Determination of the volatile fraction of coal tar

J Shanghai Univ(Engl Ed), 2010, 14(5): 313-321Digital Object Identifier(DOI): 10.1007/ s11741-010-0651-3Determination of the volatile fraction of coal tarwu Yue-ying(吴岳英)}, Jie-jun(李洁君), XU Jie(徐杰}, JIAO Chen-jia(焦展佳),WANG De-qing(王德庆)2, DING Wei- zhong(丁伟中)21. School of Environmental and Chemical Engineering, Shanghai University, Shanghai 200444, P. R. China2. School of Material Science and Engineering, Shanghai University, Shanghai 200072, P. R. China(Communicated by WU Ming-hong)SHanghai University and Springer-Verlag Berlin Heidelberg 2010Abstract The coal tar was qualitative and quantitative analyzed by gas chromatography (GC)method. 74 components wereidentified exactly by gas chromatography-mass spectrometry (GC-MS). 31 components (37%)were quantitatively analyzedby gas chromatography-flame ionization detector(GC-FID). The linearity, accuracy, precision, limit of detection(LOD)andlimit of quantitative(LoQ) determination were inspected. The scope of quantitative analysis by GC was discussed. Theexperimental results of thermogravimetric analysis(TGA)proved that gc quantitative analysis of the coal tar was reliable.Keywords coal tar, gas chromatography-flame ionization detector(GC-FID), gas chromatography-mass spectrometry(GCMS), determination, volatile fractionIntroductiona formidable challenge for virtually all important tech-niques of organic analysis. " Classical"methods(fracCoal tar, a by-product of the coking process, a sig- tional distillation(8l, column chromatography 4, 9, etc.nificant raw material in the manufacture of asphalt played the significant role in the earlier compositionalcoke, carbon black, graphitizable carbon/electrodes, studies. With the gradual improvement of chromato-and other high-value chemicals and materials such as graphic separation methods, coal tar was partitioned toindustrial naphthalene, phenol, or anthracene oil. Fur- several fractions for further analysis. Gas chromatog-thermore,release of the hydrogen in tar as H21-3) also raphy(GC)has been widely demonstrated to resolvebecomes a high-potential source of energy as an alter- numerous components of extremely complex mixturesnative to petroleum which is increasingly scarceand becomes a well-known powerful technique for analCoal tar contains a wide range of compoundsysis on the composition of coal tar(4, 5, 7-8 10-12. Thewith different molecular weights, polarities and func- flame ionization detector(FID) has high sensitivity andtionalities. Among them the most important com- selectivity for carbon-containing compounds[13], and itspounds in coal tar are polycyclic aromatic hydrocarbons wide range of linearity gives an enormous boost to thePAHs)4-8.Chemical compositions, physical proper- use of GC in coal tar analysis(7,10,14-17. Spectrometricties and structures of coal tar are not known with any techniques, such as fourier transform infrared spectromdegree of certainty. However, it is obvious that their be- etry(FTIR), nuclear magnetic resonance spectroscopyhavior, reactivity and properties are governed by their(NMR), and mass spectrometry(MS)are useful toolscomposition, and it is generally accepted that the search to characterize the coal tar or similar substances [18-21for relationships between composition, properties and Moreover, combining mass spectrometer with GC yieldsbehavior in the study of carbonaceous materials consti- more specific information of the coal tar/22-24).Thetutes the key for a more rational use of these materials. studies have reported in the application of GC methodsThe use of coal tar fractions as chemical feedstock make on the volatile fraction of coal tar, but the reliability andadditional and more detailed studies of their composi- the scope of GC method in coal tar analysis are seldomThe apparent complexity of coal tar has presented discussedtion desirableReceived Nov 25, 2009; Revised Apr 27, 2010中国煤化工Project supported by the National High-Technology Research and deveCN MHG6AAllAlKey Project of Science and Technology Commission of Shanghai(GrantCorresponding author WU Yue-ying, Ph D, Prof, E-mail: wuyy@staff. shu. edu. cr314J Shanghai Univ(Engl Ed), 2010, 14 (5): 313-321In the present study, reliable, efficient methods were 1.3 GC conditionsproposed on qualitative and quantitative analysis of the 1.3. 1 GC-MS for qualitative analysescomponents of the volatile fraction of the coal tar, withThe gC analyses of coal tar were performed withthe application by GC-MS and GC-FID. The reliability Agilent 6890N GC-5975MSD(Agilent Technologiesof the method was discussed through the data on the PaloAlto, USA)equipped with a quadrupole mass speclinearity, reproducibility and recovery. The GC feasibil- trometer with electron ionization(ionization energy ofity to the components in coal tar was also discussed by 70 eV) and a 7863B autosampler. A DB-5MS columncomparing the results between GC and thermal gravity (30 mx0. 25 mm i d, 0.25 um film thickness)connecteddirectly to the ion source and helium as carrier gas ata fow rate of 1.0 mL/ min were used. The oven tem-1 Experimentrammed from70°cto150°C1.1 Reagents and materials4°C/min,to2l8°cat2°c/min, and to300°Cat4C/min, then held at 300C for 12 min. The injecCoal tar, removed from benzene and toluene, was tor and detector temperature were set at 300C andsupplied by the Suzhou Steelworks, Suzhou, China. 200.C, respectively. The injection volume was 2.0 HLAnalytical grade carbon bisulfide was obtained from Mass spectra were recorder over the m/z range fromSinopharm Chemical Reagent Co. Ltd Impact organic 50 amu to 700 amuphase filters with a 0.20 um particle size were purchased 1.3.2 GC-FID for quantitative analysesA 0.200 g coal tar was dissolved in 30 mL carbon 1.4 Thermogravimetric analyte se %?from Shanghai ANPEL Scientific Instrument Co LtdQuantitative GC analyses of coal tar were performedAll reference materials(>95% purity) were availablewith an Agilent 6890N GC-5975MSD equipped with ancommercially from Sigma-Aldrich.FID. The same column and GC conditions1.2 Sample preparationabove and 300C FID temperature werebisulfide. The sample solution was filtrated through aThe thermogravimetrie analysis(TGa)was carried0.2 um impact organic phase filter. The residue was by an NETZSCH simultaneous TG-DTA apparatus STAwashed with carbon bisulfide, then filtrated for three 449C/ 4/ G Jupiter in pure N2 atmosphere. Temperaturetimes. All three filtrate fractions were collected into rose at 5C/min within a range from 25C to 800C.50 mL volumetric flask and diluted to mark with car- Relatively small amounts of test samples(10-15 mg)bon bisulfidewere used523134中国煤化工8000Fig 1 Chromatogram of theCNMHGJ Shanghai Univ(Engl Ed), 2010, 14(5 ): 313-3213152 Results and discussion15 peaks are not recognized because of poor separation,and 74 components are exactly identified based on a pos-2.1 Qualitative analysisitive match of mass spectral data with the NIST massThe chromatogram of the volatile fraction of coal tarspectra database(MSD)and on the comparison of GCis shown in Fig. 1. 174 peaks are obtained. Among themTable 1 Qualitative results of the coal tarRetention time/minMolecular weightComponentsp- and o-Xylenea123456783456Unknown116Benzene, 1- propynyl-677116o12Phenol, -dimethyl-aIndene, methyl-13010.4991411934Quinoline56r8g01213.582ndola1314255142Naphthalene, 1-methyl-17206156Naphthalene, 2, 7-and-2, 6- dimethyl.a, bNaphthalene, 1, 7-dimethy17704156Naphthalene, -1, 3-dimethyl-bNaphthalene, ethenyl-18202156Naphthalene,-1, 5-dimethyl-b19476154Acenaphthene197695678203233567890423420.245Naphthalenol20.655153 isomer20948170naphthalene, trimethyl-aNaphthalene, trimethyl-21.98722.390Fluorene-9-methanola22.623.254166 isomer24.06024.360182中国煤化工uran, - methyl-24.609CNMHGuran methyla26.651180Fluorene, methyl-a316JShanghaiUniv(Engl Ed),2010,14(5):313-321continuePeak NoRetention time/minMolecular weightFluorene, methyl-Methanone, diphenyl-hydrazonea67890豇23M5678002860629902178Phenanthrene30269Anthracene184Naphtho( 2, 3-b]thiophene33351Dibenzothiophene, methyl-aPhenanthrene/Anthracene, methyl-34486192Phenanthrene/Anthracene,- methyl-a350281904H-Cyclopenta(defphenanthrenePhenanthrene/Anthracene,methyl-35.401192Phenanthrene/ Anthracene methyl-aCarbazole, - methyl-aPhenathrene/anthracene, dimethyl-a39128Phenathreneanthracene, dimethyl-71Phenathrene/anthracene, dimethyl-Phenathrene/ anthracene dimethyl-Cyclic octatomic sulfur40.402Fluoranthene5678941.6549Anthracenecarbonitrilea41.962202 isomer4223342672202 isomer43.221Py43.448Benzo[b naphtho[ 2, 3-d furan, C1288588889123944.056218Benzo(b naphtho(2, 1-d furan c45.57Ben kl] xanthBenzo[b]naphtho(1, 2-d furan,45.960Pyrene/fluoranthene methyl-46.208216Pyrene/fluoranthene methyl-a46.758216Pyrene/fluoranthene methyl-46.882Pyrene/fluoranthene, methyl-47.087uyH中国煤化工 benzol fluorene47.863CNMHGBenzolbjfuoreneb48.119217Benzo[b]carbazoleJ Shanghai Univ(Engl Ed), 2010, 14(5): 313-32317continuePeak NoRetention time/minMolecular weightComponents99890mng48.793Pyrene/fluoranthene, methyl-49.13751773Pyrene/fluoranthene dimethyl-aPyrene/fluoranthene, dimethyl-5386700468g4Pyrene/fluoranthene, dimethyl-Cyclopenta(c pyrene, cBenzo(phenanthrene54453Benz( acridineBenzo(b]naphtho(2, 3-d]thiophene,c10755.8304H-Cyclopentaldd pyreneBenz a]anthracene56598TriphenyleneBenzocarbazole/217isomerUnknown57543230 isomerNaphtho 2, 1, 8, 7-klmn]xanthene5843217 isomer175930594175(4H)-Thebenidinone119595259761217 isomer, -methyl-12160010217 isomer, - methyl-a122217 isomer methyl-a60457Unknown60691Unknown61.101Binaphthalene isomer61592254 isomer64513Benzo[b Fluoranthene , c65041mzo( Fluoranthene c66014Benzo( pyrene6633766.7691M8s67017Unknown67.12067669678266 isomer6804368226H中国煤化工266 isomer&CNMHG266 isomer1446868266 isomer&69492264 isomerJ Shanghai Univ(Engl Ed), 2010, 14(5): 313-321continuePeak NoMolecular weight7093578 isomer1487109614971213Dinaphtho[ 1, 2-b: 1, 2-d thiopheneBenzo[b triphenyleneIndeno[ 1, 2, 3-cd fluorancene/Indeno[1, 2, 3-cd pyrene,72.216Dibenz[a, hJanthracenec27815572633278Benzo b]chrysene15773.095276Benzo(ghi] peryleneDibenzoldef, mno]chrysene15973.834Unknown74.749nzoperylene-methyl-a16174290Benzoperylene-methyl-a1678.103302 isomer16578.601302 isomer16779106302302 isomer79709302 isomer80095300CoroneneNaphtho-/ Dibenzo-pyrene isomer&,c17180417302Naphtho-/ Dibenzo-pyrene isomer, c7280819Naphtho-/ Dibenzo-pyrene isomer80988302Naphtho-/Dibenzo-pyrene isomer, c17481391aphtha/Dibenzo-pyrene isomer, cNote: Identified by matching of standard MSD; Identified by comparing the retention time of reference materials;Identifiedaccording to the retention data from reference literatures 8, 15, 25retention time of reference materials and jor components of the coal tar. For most of themliteratures,11,251.85 peaks can not be exactly identified the quantitatively analytical results are believable, be-because they are isomers and difficult to distinguish by cause of good linearity(r>0.999)in calibration curves,MSD. The GC retention indexes and molecular weight precision(1.0%0-8.2%), and accuracy(recovery factorsinformation for the compound assigned to each peaks 88%-116%). The highest content reaches 103.2 mg/gin Fig. l are summarized in Table 1(tar) for naphthalene. The total quantity of 32 com-2.2 Quantitative analysisonents are 370. 4 mg/g(37%)in coal tar and took68.5% in total peak area. It was estimated that aboutInjector temperature and final oven temperature are 54% of sample were possible to be quantitatively an-set at 300C. for complete gasification and elution ofthe components with high boiling point(BP). The op-alyzed by GC. However, thee components with abouttimum GC conditions described in Section 1.3.2 are the 31.5% in total peak area of the GC-FID chromatogramsame as those applied in GC-MS qualitative analysis. are not quantitatively analyzed because their standardThe flow rates of carrier gas(He), fuel gas(H2)and air materials were not available or their peak areas wereare selected by a series of investigations to have high中国煤化工ely. Some peaks insensitivity and stability for the components to be quan.thenot presented in thetitatively analyzedCN MH Ge sensitivity of FIDTable 2 summarizes quantitative results of the ma-was not as high as that of MS. For quantitative data, aJ Shanghai Univ(Engl Ed), 2010, 14(5): 313-321319Table 2 Quantification results of the major components of the coal tarCorrelation Area/ Contentab/ Recovery/ BP/ LoD/ LoQ/Componentscoefficient r%(mg g-)(ug.g"")(ug.g")Indene099982.41012501.2097218267.42Naphthalene0999820.200103.205.60113,0218.08.900.99970.4601000Isoquinoline099980.1120477.40116.00.999803383.17Naphthalene, 2-methyl-1.0000246012.503.80101.02413.366.720.9997244.63.90Biphenyl0.99994.414.20254.31.222.44099940.200Naphthalene,- 1, 7-dimethyl-0.99960.2423.3749088.7Naphthalene,- 1, 3-dimethyl-0.999803811.84965Naphthalene, 1, 5-dimethyl-099960.24788.624.200.99954.4622305.00929Acenaphthene0.999039499221.40Dibenzofuran0.99961,64010.601.0000281014804.101.78099950.4902193.101020099997.8604.801.793.58Anthracene09995264013902.20Acridine0725709913462014.004H-Cyclopenta def phenanthren5686.401020N/AI134Fluoranthene28704.100.99953.700105011H-Benzo(a fluorene0.99980.890134.4054811H-Benzo[b fluorene09998103.6942enza anthracene8261.007.03Triphenylene9477.80126009999063130Benzole pyrene0.99990990106.027.40Benzo(apyrene099969586Perylene0.99990452355720924350-40029000.99720.7550013.5277551.200.9953N/AEN/A3704Other components31.500170.3°by br and 10 gained f ram of coal tar sample; b The relative standard deviation(RSD )obtained in five determinationsNote:The unit is mgperThe recovery olom the average of three determinations; The LOD and the loQ were calculated as Ci revealedrespectively, where Ci was the concentration of the component i through the linear equations obtainedusing external standard method and N is the average background noise in the range of corresponding retention time in tendeterminations of blank samples; Estimated by the normalization method; Not detected10o difference is required. Limits of detection(LODs) of benzo ghi]perylene and coronene were not as goodand limits of quantitative determination(LoQs)are as other components. A relatively poor linearity(rassigned when the signal of the component exceeded 0.9972), precision(RSD= 13%)and accuracy(recovthe mean value of the background noise of 5o and 10o ery fa中国煤化工 r benzolghi perylenerespectively. The lower sensitivity of some components Therewith high BP could be due to insufficient vaporization. coroCNMHGhether a componentIt is notable that the quantitatively analytical results is determined by GC mainly depended on its BP. TheJ Shanghai Univ(Engl Ed), 2010, 14(5): 313-321behaviors of the components with high BP in GC are 2.3 Thermogravimetric analysisstrongly affected with the variation of GC apparatusIn order to study the volatility of the coal tar, a tGaand the analytical results became unstableis undertaken. The results in Fig. 2 show the relationshipThe quantitative results of benzole pyrene, between sample weight losses and testing temperaturebenzo( a pyrene and perylene are acceptable but those of It means that the volatile fraction in the coal tar in-benzo ghi perylene and coronene are not. Therefore we creases with temperature going up. It is noted that aconsider that the component with BP lower than 495C volatile fraction are 50%at 300C, the same temperin the coal tar could be quantitatively analyzed well ature as that of injector and final column oven in gCby gCanalysis. The agreement between the result estimatedMass change:-422Mass change:-5004Fig2 TGA result of the coal tarfrom GC-FID chromatogram and that from TGa proves Referencesthat the quantitative analysis on the coal tar by gc isreliable. about half quantity of the sample could not be1 ONOZAKI M, WATANABE K, HASHIMOTO T, SAEGUSAKATAYAMA Y. Hydrogen preanalyzed by gC methooxidation and steam reforming of tar from hot coke ovengasJ]-Fuel,2006,85(2):1431493 Conclusions2 SIMELL P A, HEPOLA J O, KRAUSE A O. Effects ofgasification gas components on tar and ammonia de-composition over hot gas cleanup catalysts J. Fuel,)174 peaks are obtained in GC-MS chromatogram1997,76(12):111-112from coal tar sample. Among them 74 components are3 SIMELL P A, LEPPALAHTI J, BREDENBERG J B Cat-exactly identified but 15 peaks were not recognizedalytic purification of tarry fuel gas with carbonate rocksOther 85 components are not identified exactly by GCand ferrous materials J]. Fuel, 1992, 71(2): 211-218MS14 BORWTTZKY H, SCHOMBURG G Separation and iden-ii)31 major components in the coal tar sampletification of polynuclear aromatic compounds in coal(37%)are quantitatively analyzed by GC-FID with goodtar by using glass capillary chromatography includingaccuracy and precision. About 54% of the coal tar werecombined gas chromatography-mass spectrometry Jpossible to be quantitatively analyzed by GC. It is foundJournal of Chromatography A, 1979, 170(1 ):99-124that the components with BP lower than 495C car( 5] NOVoTNY M, STRAND J W, SMITH S L, WIESLER D,quantitatively analyzed well by GC methodI, Comnooitinnal studies of coal tar byAn experiment on TGA is undertaken. The agree-中国煤化工 mass spectrometryment between the results by GC and that by TGACNMHGproves that the GC quantitative analysis on the coal [6]1. Cu uue wulyuoiiun of pitches [J]. Fuel,tar is reliable1987,66(11:1536-1539J Shanghai Univ(Engl Ed), 2010, 14(5): 313-321321[7 DOMINGUEZ A, ALVAREZ R, BLANCO G C, DIEZ M [16] GUILLEN M D, DOMINGUEZ A, LGLESIAS M J, BLANCOA Chromatographic evaluation of some selected poly-C G. Semiquantitative gas chromatographic analysisclic aromatic hydrocarbons of coal tars produced un-of the volatile fraction in several extracts obtained byder different coking conditions and pitches derived fromtreatment of coal tar pitches with different organic sothem [J. Journal of Chromatography A, 1996, 719(vents同J].Fue,1995,74(2):233-240181-194.[17] WISE S A, BENNER B A, BYRD G D, CHESLER S N,[8zhangMJ,ChenBJ,SheNSD,cHenSy.com-REBBERT R E, SCHANTZ MM. Determination of polypositional studies of high-temperature coal tar by g ccyclic aromatic hydrocarbons in a coal tar standardFT-ir. analysis of middle oil fractions [J]. Fuel, 199reference material [J]. Analytical Chemistry, 199876(5):41542360(9):887-89 LIYA E Y, LYNN M H, JoHN D, STEPHEN N Charac(18 SUN X G. The investigation of chemical structure ofterization of coal tar organics via gravLmncoal macerals via transmitted-light FT-IR microspec-troscopy J]. Spectrochimica Acta Part A: Molecularand Biomolecular Spectroscopy, 2005, 62(1-3):557-[10] GUILLEN M D, IGLESIAS M J, DOMINGUEZ A, BLANCOC G. Polynuclear aromatic hydrocarbon retentiondices on SE-54 stationary phase of the volatile comp[19 CANTOR D M. Nuclear magnetic resonance spectromet-nents of a coal tar pitch: Relationships between chro-ric determination of average molecular structure pamatographic retention and thermal reactivity J]. Jour-rameters of coal-derived liquids J. Analytical Chem-al of Chromatography A, 1992, 591(1-2):287-295Istry,1978,50(8):1185-1187H R. Coal[11] BLANCO C G, BLANCO J, BERNAD P, GUILLEN Mtar analysis by mass spectrometry -a comparison ofD, Capillary gas chromatographic and combined gaschromatography-mass spectrometric study of themethods.Fue,1993,72(1):31-43olatile fraction of a coal tar pitch using ov-1701 sta21 LDZARo M J, HEROD AA, CoCKSEDGE M, DOMINtionary phase J]. Journal of Chromatography A,1991M. KANDIYOTi R. Molecular mass determinations in539(1):157-167.coal-derived liquids by MALDI mass spectrometry andsize-exclusion chromatography J]. Fuel, 1997, 76(13)2] ALCANIZ-MONGE J, CAZORLA.AMORO D, LINARES-1225-1233.SOLANo A. Charactersation of coal tar pitches by thmal analysis, infrared spectroscopy and solvent frac-22 JoHN R K, BLACK K J T Structural characterizatiotionation [ J]. Fuel, 2001, 80(1 ) 41-48of coal-tar and petroleum pitches [J]. Energy and Fuels,1993,7(3):420425.[13] ScHOFIELD K. The enigmatic mechanism of the fame (23)ALVAREZ R, DIEZM A, GARCIA R, DE G, ANDRE Aionization detector: Its overlooked implications for fos-L, SNAPE C E, MOINELO S R. Effect of preheating insil fuel combustion modeling J. Progress in Energycarbonization on the nature of resultant coal tars andand Combustion Science, 2008, 34(3): 330-350pitches J]. Energy and Fuels, 1993, 7(6): 953-959[14] BLANCO C G, CANGA J S, DOMINGUEZ A, IGLESIAS (24 MARTIN Y, GARCIA R, SoLE R A, MOINELO SR Struc-M J. Flame ionization detection relative response fac-ture characterization of coal tar pitches obtained bytors of some polycyclic aromatic compounds: determi-heat treatment under different conditions J. Energynation of the main components of the coal tar pitchvolatile fraction J] Journal of Chromatography A1992,607(2):295302(25] WANG Z D, LI K, LAMBERT P, YANG C. Identification,characterization and quantitation of pyrogenic pol[15 DOMINGUEZ A, BLANCO C, SANTAMARIA R, GRANDAcylic aromatic hydrocarbons and other organic com-M, BLANcO C G, MENENDEZ R. Monitoring coal-tarpounds in tire fire products J. Journal of Chromatog-pitch composition changes during air-blowing by gasraphy A,2007,1139(1):14-26.hromatography J. Journal of Chromatography A,2004,1026(1-2):231-238(Editor JIANG Chun-ming)中国煤化工CNMHG