Determination of constituents of essential oil from Angelica sinensis by gas chromatography-mass spe

Vol. 12 No. 4J CENT. SOUTH UNIV. TECHNOL.Aug.2005Article id:1005-9784(2005)4-0430-07Determination of constituents of essential oil from Angelica sinensis bgas chromatography-mass spectrometryWU Ming-jian(吴名剑), SUN Xian-jun(孙贤军), DAI Yuan-hu(戴云辉)GUO Fang-qiu(郭方遒), HUANG Lan-fang黄兰芳), LIANG Yi-en梁逸曾(1. Technical Center, Changde Cigarette Factory, Changde 415000, China i2. School of Chemistry and Chemical Engineering, Central South University Changsha 410083ChinaAbstract Gas chromatography -mass spectrometry( GC-MS )coupled with chemometric resolution upon two-dimensionaldata was employed to analyze the constituents of essential oils of Angelica sinensis. Constituents in essential oils of Angelicasinensis root were identified by GC-MS with the help of subwindow factor analysis( SFA method resolving two-dimensionaloriginal data into mass spectra and chromatograms. 76 of 97 separated constituents in essential oil of Angelica sinensis rootwere identified and quantified, and they account for about 91. 36% of the total content. The results show that ligustilidebutylene phthalide 2-methoxy 4-vinylphenol carvacrol allo-ocimene 2 6 6-trimethylbicyclo-f3 1 Jhept-2-ene are themain constituents in essential oil of Angelica sinensis rootKey words: Angelica sinensis chemometrics gas chromatography-mass spectrometry subwindow factor analysisCLC number: 0657.7*1. R284Document code: A1 INTRODUCTIONused for tonifying the blood and for treatment of anemia, rheumatism, female menstrual disorders and amelTraditional Chinese medicines( TCMs )are very orrhea. The essential oils are main pharmacological acimportant due to their extensive use in preventing and tive components of Angelica sinensis. Much work hascuring many diseases for humans. Although only some been done upon its essential oils 12-15.However, onlycompounds might be responsible for pharmaceutical a small number of constituents were identified due toeffects, traditional Chinese medicines usually contain inevitable overlapping-peaks in such a complicated systoo many components and form a very complex system, tem despite the optimal separationwhich makes the screening and analysis of bioactiveIn this paper, the essential oils of Angelica sinen-components extremely difficult I compared with those sis were studied with GC-MS, followed by the resoluof western medicines. When such complex samples are tion of the two-dimensional data with subwindow factoranalyzed, it is common that different contents of multi- analysis( SFA )method 0 J. Mass spectra of eachcomponents and great differentiation in the complex component can be obtained through unique resolutionsystems make the complete separation and exact detec- Then, qualitative identification of these constituentslytical instruments axe employed ood andtion rather difficult even though good andaccurate anawas performed using retention time and mass spectrathe overall volume integration method s4, tied out withFinally the quantitative analyses were carRecently with the advance of hyphenated instru-ments, multidimensional data revealing the compositions of samples can be obtained from GC-MS, HPLC- 2 THEORYDAD and so on 2-4]. Then, many associated methodshave been developed to provide more information forHere theoretical explanation is showed conciselychemical analysis both in chromatographic separation for the sake of brevity[ 10 d1 1and in spectral identification which makes it possibleAccording to the lambert-Beer Law a two-dimen-o interpret these complex systems 5-11.sional data Xmxn produced by GC-Ms can be expresseAngelica sinensis( Oliv. Diels( Umbelliferaenamed Dong quai a very important herb in China中国煤化工CNMHGo Foundation item: Projec( 20175036, 20235020)supported by the National Natural Science Foundation of China projec([ 2004 ]14)supportedby Foundation of Hunan Development and Innovation CommitteeRe吗点RUANG Lan-fang, PhD; Tel: +86-731-8836376: E-mail If1 8 huangwU Ming-jian , et al: Determination of constituents of essential oil from Angelica sinensisas followswith a fused-silica capillary column( 30 m x0. 25 mmX=CS+Ei. d. )coated with Ov-1 methyl siliconeXmxn=2C S+E(=l2 m.N)(1) phase was used to obtain chromatograms of the fracwhere Xnxn denotes an absorbance matrix represen- tion. The oven was held at 70C for I min during inting N components of m chromatographic scan points atjection then temperature programmed at 2. 5C/minn atom mass units or wavelength points. Cpoints.C is the pure to a final temperature of 250C and held for 20 minhromatographic matrix ,S is the pure mass spectral Inlet temperature was kept at 290 C all the time. 0.04matrix and E represents the noise. The unique resolution of a two-dimensional data into chromatograms anduL essential oil was injected into the gC. Helium carspectra of the pure chemical constituents is obtainerier gas at a constant flow-rate of 0. 5 mL/ min and awith SFA method. The whole procedure goes in the fol- split ratio of 6: I were used simultaneously. A Shilowing stemadzu QP-5000 mass spectrometer was operated in full1)Pretreat the original matrix. The measured ma- scan and electron impact( EI +)modes with an electrix is divided into different submatrices by zero compo- tron energy of 70 eV. In the scanning range of 35 toThen subtract the background out using 450 m/z, mass spectra were recorded with 0method based on principal component analysi(PCA2)Establish a rank map and determine number of3.5 Data analysespure components in the target cluster ;Data analyses were performed on a Pentium basedindicated by the rank map with the help of S?. onent IBM compatible personal computer. All programs of3)Esthe chemometric resolution methods were coded in4)Verify the reliability of the resolved results MATLAB 5. 3 for windows. The library searches and5 )Qualitative analysis is performed by similarity spectral matching of the resolved pure components weresearches in the National Institute of Standards andconducted on the nist ms database containing aboutTechnology( NIST)mass database and quantitative re-107 000 compoundssults are obtained by calculating the volume of totaltwo-wav response4 RESULTS AND DISCUSSION3 EXPERIMENTAI4. 1 Component resoluti3.1 InstrumentsThe total ionic current( TiC chromatogram of es-A GC-17A gas chromatograph and QP-5000 mass sential oil of main root of Angelica sinensis is shown inspectrometer hyphenated instrument from Shimadzu Fig. 1. It can be seen that there are a lot of peaks andwere etheir contents vary greatly. Although chromatographic3.2 Materialsverla with one another which mayAngelica sinensis from Gansu Province of China was rect identification of compounds without further datapurchased from Jiuzhitang Pharmacy( Changsha, Chi- processing. For example, different qualitative resultscan be obtained at different positions of one peak if it is3.3 Extraction of essential oilcomponent with low content is very difficult to be iden-All these raw herbs were dried at 40C for abouttified straight with the nist mass database, since two40 min. Some 400 g dried Angelica sinensis and 1. 2 Ldistilled water wereThey were placed into adimensional data obtained by mass spectral measurestandard extractore mixture was kept for 30 min at ment unavoidably contain peaks associated with columnroom temperature before extracting the essential oilbackground and residual gasesEssential oil was obtained by the standard extractinHowever, if the overlapped peaks and the commethod for essential oil in TCMs according to the chinents with low content are resolved into pure spectranese Pharmacopoeia 6 The extract from root of AnarV凵中国煤化工ative analysis of compo-relica sinensis is 0. 512%( mass fractionnentCNMH Gable degree. Taking theocglicit a within 14.0- 14. 6 min3. 4 Detection of essential oilnamed peak cluster C, for example the wholeA Shimadzu 17A gas chromatograph equipped432Journal csut Vol 12 No, 4 2005In the first place it is necessary to correct back-ground and baseline shift with PCA[8. Fig. 2( a)denoted by curve shows the chromatogram after back-ground removing. Obviously baseline is down to thebottom. In the second place it is important to estimatethe eluting sequences of individual components by theuse of fix-sized moving window evolving factor analy1.0FSMWEFA )S to determine the number of pure com-0.5ponents hidden in the peak cluster after backgroundThe rankobtained from FSMWEFAshown in Fig 3, which provides a clear insight intoevery retention time and the eluting information of conFig 1 TIC chromatogram of essentialcentration profiles can be obtained. The regions laof angelibeled 1, 2, 3 and 4, respectively are possibly oncomponent regions while two or three species are coe-procedure of this approach can be demonstratedFig 2( a)denoted by curve I shows an originalluted in the regions labeled 1 +21+2+32+3 3+ 4. Finally subwindow factor analysis is applied tochromatogram from 14. 0 min to 14. 6 min( peak clus-er C). It can be seen that the frontal peak is not therectly extract component spectra from overlappingbaseline separated from the latter peak. Both frontalpeak of segment X. With the information mentionedpeak and latter peak are not a single component bebove determination of both left and right subwindowsof some components for the use of SFA also becomescause different spectra are obtained at different poeasy. And to Sfa method, its ability to examine andtions. By direct search from mass library matchingesolve the target data does not depend excessively onalues are quite low lower than 65%). From the curthe selection of the window size. In order to extract therent results the quantitative analysis of this peak clus-er seems also impossible, because the area of eachsubwindow is chosen from 4 230 to 4 250 scan pointscomponent Is unknownfrom 14 10 min to 14 17 min )and the right one from4 260 to 4 310 scan points( from 14 20 min to 14. 37min ) The first largest singular value d, is 0. 98 whereas the second one is 0. 14, i.e. d2being close to 0. Thus, the spectrum of component 1denoted by s, can be acquired. In this way the spectraof others in peak cluster C are got finally. These re-4.014.214414.6solved spectra constitute spectral matrix S. If all thet/minpure spectra are available the concentration profilescan be achieved by using prior information of spectraand linear regressionC=XS(SS)Thereby their concentration profiles could beectra and Egn(2), which are marked with number 1, 2, 3 and 4, in4.014.214414.6turn in Fig. 2(b)42rV凵中国煤化工Fig2 TIC chromatogram of peak clusterand resolved chromatographic profilesolved catograpnlc proteles of these four compo-of peak cluster C b)nents are obtained, identification of them can be doneby similarity searches in the NIST mass database andI-Chromatogram after background removin方数据wU Ming-jian , et al: Determination of constituents of essential oil from Angelica sinensis4331346.0Jhn1+213380134l05114.214.314.4.514.6t/min100150Fig 3 Evolving eigenvalues obtained usinmzFSMWEAF with a window size of 8 of peak Cverified with retention indices. Components 1 2 and 4Fig 5 Resolved mass spectra of component 2by SFAC a and standard mass spectrum( b )ofmay be 2 A-dimethy benzaldehyde 3 A-dimethylbenz3 A-dimethylbenzaldehyde( Cg hoo( 134 ))aldehyde and[-]-4-terpineol in order with the matchseeFigs 4 5 and 6 respectively ) Component 3 is not i-dentified due to the very low matchof0.6241435617111150m/z80120160200120160200Fig 6 Resolved mass spectra of componentFig 4 Resolved mass spectra of component4 by SFAC a )and standard mass spectru( b)I by SFAC a )and standard mass spectrum( b )ofof[-]4-terpineol( C1o,8 O 154))2 A-dimethylbenzaldehyde(Cg hoo( 134 ))spectrum of each component are resolved, the totalIn the same way the spectra of components itwo-way response of each component can be obtainedfrom the outer product of the concentration vector andother segments can also be obtained. The qualitative the spectrum vector for each component. The total aresults are listed in Table 1, where B is relative contentmount of each component is proportional to the overallof identified component. 76 of 97 separated constitu-The advantage of thisents in essential oil of main root were identified com中国煤化工paring this result with those of literature[ 12-141, this quancHal peak-area integrationCNMH Gng points are taken intocombined approach works much better and more comconsideration. The identified components representponents are identified satisfactorilyHunan tail ai e est th lald o betHe4.3 Quantitative analAfter tHe pure chromatographic profile and mass434Journal csut Vol 12 No, 4 2005Table 1 Analytical results of essential oil from Angelica Stotompound nameMolecule structure1.982ButanalChOC?H41234567894.639CaH4.928NonaneBeta pineneC10H16.359OctaneCRH,6.8107.1497-Methyl-3-methylene-1 6-octadiene10C10H18.1411-Methyl-2( 1-methylethyl )-benzene138.6822 6 6-Trimethylbicyclo-t 3 I J hept-2-eneC10H164.723 7-Dimethyl-l 3 7-octatriene4-Methyl-3 1-methylethylidene F-cyclohexene9.793UndecaneCIH678910.41010.605Trans trans-Hexa-2 A-dienyl acetateCHIo,10.668Allo ocimene10.895-DeceneCuHlL.0346 7-Dimethyl-bicyclo[ 4.2.0]octanel1.379l1.892I 5 5-Trimethyl-6-methylene-cyclohexeneC10H160.38l2.3355-Methyl-3( 1-methylethylidene )1 A-hexadiene13.509C1H560.2613.612Butyl-1 A-cycloheptadiene13.9455-Decanone4.182 A-DimethylbenzaldehydeCIo H14.2853 4-DimethylbenzaldehydeCo Ho14.476ChiSO5.321[-]-alpha- Terpineo[ P-menth-1-en-8-olC1o Hiso17.2455-Decen-1-oIC1oH20 O19.3521-Methoxy 4[ 1-propenyl ]benzeneCohO5.Methvl-2-deceneV凵中国煤化工0.x420.2524. Methvl-5-decanolCNMHGH2O20.6286-Hydroxymethy l-2 3-dimethylphenyl methanol20.9492 9-Dimethyldecane21.6142-Methoxy 4-vinylphenolCo H1 owU Ming-jian , et al: Determination of constituents of essential oil from Angelica sinensis435Continued tableRetention time/ompound nameMolecule structure22.7722 A 5-TrimethylbenzaldehydeChO24.82010-UndecenalC25.1042 5-Dimethyl-3-hexanol acetate26.0282-Methyldodecane26.270+]beta-Ced27.4275 9 9-Trimethyl-spiro[ 3 5 non-5-en-1-oneC1 His2-ButenyI hexanoic acid ester5 6-Dihydro 4[2 3-dimethyl-2-buten 4-yl ]-2H-pyran-Cuhuo5330.310CIs H0.3930.429beta-HimachaleneC15H240.34C131.0283 7 11-Trimethyldodeca-1 6, 0-trien-3-oICis H2so31.660C15H240.35[+]alpha-LorCis HMethyl[ Z 12[ 3-cyclopropyl-7-norcaranyl acetateCu HnO0.362-Hexenyl ester,[ E H-benzenacetic acid0.3435.681Dodecvl-oxirane35.8182-TetradecanolCHoo36.015Cis-7-Tetradecen-1-oIC1H28 O38.226Butylidene phthalideIso-butylicene phthalide43.194C14H14O244.543Cis H24046.748E-7-Tetradecen-1-olCu Haso0.3647.I1-Hexadecene48.144tetradecanoic acidC14H28O248.663E-10-PentadecenolCIs H3o o51.076E-10-Hexadecadiene54.03757 8-Trimethyl-dihydrocoumarinC12H14O257.342Cis-7-Tetradecen-l58.820Stearolic acidV凵中国煤化工5 CONCLUSIONSCNMHGStituents in essential oilof Angelica sinensis root are identified and quantifiedA Gas chromatography-mass spectrometry GCaccount for about 91. 36 of the total conMs )coupled with combined chemometric method, tent. This study shows that the application of combinedsuch as PCA, FSMWEFA and SFA, for analysis of approach is a powerful tool which can be used for re-constituents nxE ntial oils of Angelica sinensis is de- vealing the quality and quantity of chemical constitu-436Journal csut Vol 12 No, 4 2005ents of traditional medicines effectivelynal pmethod for two-way data resolution JAnalyst,1999,124(10):1471-1476[9] Liang Y Z, Kvalheim O M, Keller H R, et al. HeuristicREFERENCESevolving latent projections: resolving two-way multicom-onent data. Part 2: Detection and resolution of minorconstituents[J]. Anal Chem, 1992, 64(8): 946-953III Wang H, Zou H, Kong L, et al. Analysis of bioactive [10] Manne R, Shen H L, Liang Y Z. Subwindow factor a-components in traditional Chinese medicines by molecularnalysis[ J ] Chemom Intell Lab Syst, 1999, 45(1biochromatography with alphal-acid glycoprotein stationa-2):171-176ry phase[ J ] J Basic Clin Physiol Pharmacol 2000, 11 [11 Shen H L, Manne R, Xu Qs, et al. Local resolution of(2):155-172.hyphenated chromatographic data J ] Chemom Intell2 Liang Y Z. White, Gray and Black Multicomponent Sys-Lab Syst,1999,4X1-2):323-330.tems and Their Chemometric Algorithms[ M ] ChangshaHunan Science and Technology Press, 1996.( in Chi12 CHENG Yao-zu, CHENG Neng-yi, MA Xue-yi, et alStudy on the constituents of Angelica sinensis [J]neseChemical Journal of Chinese University 1984, 5(1)[ 3] Gong F, Liang Y Z, Xu Q s et al. Determination of vol-125.( in Chineseatile components in peptic powder by Gas chromatography-mass spectrometry and chemometric resolution[ J ]. J 13 CHENG Nan, XUE Dun-yuan. Study on the SC-CO, ex-Chromatography A, 2001, 909(2): 237-247tracted of the essential oil in Angelica Sinensis[J][4] Li X N Cui H, Song Y Q, et al. Analysis of the volatileJournal of Gansu Educational College( Natural Sciences),2003,17(4):30-33.( in Chinese)fraction of schisandra Chinensis( Turcz. ) baill. with [14 LI Gui-sheng, MA Cheng-jun, LIU Zhi-feng ,et al. ExGC/MS and chemometric resolution[ J ] Phytochemicaltraction of the essential oil from Angelica sinensis by su5) nalysis,200,141:23-33Keller H R, Massart D L. Peak purity control in liquidpercritical-CO, fluid in comparison with that with streamdistillation[J]. Chinese Traditional and Herbal Druhromatography with photodiode-array detection by a fixedsize moving window evolving factor analysis J J. Anal 15] Chinese Pharmacopoeia Committee. Pharmacopoeia of the[6] Maeder M, Zilian A. Evolving factor analysis,aPeople's Republic of China( English Edition ) M ]Guangzhou: Guangdong Science and Technology Pressmultivariate techniques in chromatography[ J ] Chen1988ntell Lab Syst,1988,33):205-213.[7] Maeder M, Zuberbuhler A D. The resolution of overlap- 16] Chinese Pharmacopoeia Committee. Chinese Pharmaco-ing chromatographic peaks by evolving factor analysispoeia-Appendix 62[ M ] Beijing: People's Health[J] Anal chim acta,1986,18l(2):287-291Press, 1995. in Chinese[8] Xu C JJ H, Liang Y Z. Evolving window orthogo-Edited by ZHAo Jun中国煤化工CNMHG