The Traditional Herbal Medicines

7/31/2019 The Traditional Herbal Medicines

1/10

1

The traditional herbal medicines (HM) and their preparations have beenwidely used for thousands of years in many oriental countries, such as in China,Korea, Japan, etc. However, one of the characteristics of oriental herbal medicinepreparations is that all the herbal medicines, either presenting as single herbs or ascollections of herbs in compositeformulae, are extracted with boiling water during the

decoction process. This may be the main reason why quality control of orientalherbal drugs is more difficult than that of western drug. As pointed in GeneralGuidelines for Methodologies on Research and Evaluation of Traditional Medicines(World Health Organization, 2000) [1], Despite its existence and continued useover many centuries, and its popularity and extensive use during the last decade,traditional medicine has not been officially recognized in most countries.Consequently, education, training and research in this area have not been accordeddue attention and support. The quantity and quality of the safety and efficacy data ontraditional medicine are far from sufficient to meet the criteria needed to support itsuse world-wide. The reasons for the lack of research data are due to not only tohealth care policies, but also to a lack of adequate or accepted research

methodology for evaluating traditional medicine. In general, one or two markers orpharmacologically active components in herbs and or herbal mixtures were currentlyemployed for evaluating the quality and authenticity of herbal medicines, in theidentification of the single herb or HM preparations, and in assessing the quantitativeherbal composition of an herbal product. This kind of the determination, however,does not give a complete picture of a herbal product, because multiple constituentsare usually responsible for its therapeutic effects. These multiple constituentsmaywork synergistically and could hardly be separated into active parts. Moreover, the chemical constituents in component herbs in the HM products may varydepending on harvest seasons, plant origins, drying processes and other factors.Thus, it seems to be necessary to determine most of the phytochemical constituentsof herbal products in order to ensure the reliability and repeatability ofpharmacological and clinical research, to understand their bioactivities and possibleside effects of active compounds and to enhance product quality control [24]. Thus,several chromatographic techniques, such as high-performance liquidchromatography (HPLC), gas chromatography (GC), capillary electrophoresis (CE)and thin layer chromatography (TLC), can be applied for this kind of documentation.In this way, the full herbal product could be regarded as the active compound. Theconcept of phytoequivalence was developed in Germany in order to ensureconsistency of herbal products [5]. According to this concept, a chemical profile,such as a chromatographic fingerprint, for a herbal product should be constructed

and compared with the profile of a clinically proven reference product.In 2004, the Chinese State Food and Drug Administration (SFDA) will regulate thecompositions of liquid injection with HM ingredients using stringent qualityprocedures such as chemical assay and standardization. Fingerprints of HM liquidinjections are compulsorily carried out for this purpose. In addition, among thevarious experimental techniques, chromatographic methods are highlyrecommended for finding out the fingerprints of these products [612].By definition, a chromatographic fingerprint of a HM is, in practice,a chromatographicpattern of the extract of some common chemical components of pharmacologicallyactive and or chemically characteristics [67,11]. This chromatographic profileshould be featured by the fundamental attributions of integrity and fuzziness or

sameness and differences so as to chemically represent the HM investigated[11,13]. It is suggested that with the help of chromatographic fingerprints obtained,


2/10

2

the authentication and identification of herbal medicines can be accuratelyconducted (integrity) even if the amount and/or concentration of the chemicallycharacteristic constituents are not exactly the same for different samples of this HM(hence, fuzziness) or, the chromatographic fingerprints could demonstrate both thesameness and differences between

various samples successfully [11,14]. Thus, we should globally consider multipleconstituents in the HM extracts, and not individually consider only one and/or twomarker components for evaluating the quality of the HM products. However, in anyHMand its extract, there are hundreds of unknown components and many of themare in low amount. Moreover, there usually exists variability within the same herbalmaterials [4,15]. Consequently, to obtain reliable chromatographic fingerprints thatrepresent pharmacologically active and chemically characteristic components is notan easy or trivial work. Fortunately, chromatography offers very powerful separationability, such that the complex chemical components in HM extracts can be separatedinto many relatively simple sub-fractions. Furthermore, the recent approaches ofapplying hyphenated chromatography and spectrometry such as high-performance

liquid chromatographydiode array detection (HPLCDAD), gas chromatographymass spectroscopy (GCMS), capillary electrophoresis-diode array detection (CE-DAD), HPLCMS and HPLCNMR, could provide the additional spectral information,which will be very helpful for the qualitative analysis and even for the on-linestructural elucidation.With the help of the spectral information the hyphenatedinstruments show greatly improved performances in terms of the elimination ofinstrumental interferences, retention time shift correction, selectivity,chromatographic separation abilities, measurement precision [1618]. If hyphenatedchromatography is further combined with chemometric approaches, clear picturesmight be developed for chromatographic fingerprints obtained. These excellentproperties are the so-called dimension advantages proposed by Booksh andKowalski [19]. A chemical fingerprint obtained by hyphenated chromatography, out ofquestion, will become the primary tool for quality control of herbal medicines [1315,2037].2. Chromatography and chemical fingerprints ofherbal medicinesIn general, the methods for quality control of herbal medicines involve sensoryinspection (macroscopic and microscopic examinations) and analytical inspectionusing instrumental techniques such as thin layer chromatography, HPLC, GCMS,LCMS, near infrared (NIR), and spectrophotometer, etc. [38]. On the other hand,the methods of extraction and sample preparation are also of great importance in

preparing good fingerprints of herbal medicines. In this review article, however, weshall only focus our attention on how to construct reasonably the chromatographicfingerprints and their reasonable and efficient evaluation for the purpose of qualitycontrol. As a single herbal medicine may contain a great many natural constituents,and a combinationof several herbs might give rise to interactions with hundreds ofnatural constituents during the preparation of extracts, the fingerprints produced bythe chromatographic instruments, which may present a relatively good integralrepresentation of various chemical components of herbal medicines, are mainlyconcerned in this review.2.1. Thin layer chromatographyTLC was the common method of choice for herbal analysis before instrumental

chromatography methods like GC and HPLC were established. Even nowadays,TLC is still frequently used for the analysis of herbal medicines since various


3/10

3

pharmacopoeias such as American Herbal Pharmacopoeia (AHP) (Upton, SantaCruz, US, 2002), Chinese drug monographs and analysis (Wagner, Kotzting/Bayer,Wald, Germany, 1997), Pharmacopoeia of the Peoples Republic of China (ChemicalIndustry Press, Beijing, 1997), etc. Still use TLC to provide first characteristicfingerprints of herbs.

Rather, TLC is used as an easier method of initial screening with a semiquantitativeevaluation together with other chromatographic techniques. As there is relatively lesschange in the simple TLC separation of herbal medicines than with instrumentalchromatography, only a brief summary is given here, and for further details aboutTLC the readers could consult references [3940].TLC has the advantages of many-fold possibilities of detection in analyzing herbalmedicines. In addition, TLC is rather simple and can be employed for multiplesample analysis. For each plate, more than 30 spots of samplescan be studied simultaneously in one time. Thus, the use of TLC to analyze theherbal medicines is still popular [4150,221224]. With the help of the CAMAG videostore system (CAMAG, Switerland) and TLCQA-UV methods[51], it is possible to get

useful qualitative and quantitative information from the developed TLC plate. Onecan observe that, in reference [51], the four samples of Cordycepssinensisfrom thejoint products of China and Japan cooperation have more valuable medical effectcompared to others as they contained the most effective component cordycepin.Moreover, with the help of image analysis and digitized technique developed incomputer science, the evaluation of similarity between different samples is alsopossible. In summary, the advantages of using TLC to construct the fingerprints ofherbal medicines are its simplicity, versatility, high velocity, specific sensitivity andsimple sample preparation. Thus, TLC is a convenient method of determining thequality and possible adulteration of herbal products. It is worth noting that thetechnique of TLC are also being updated in progress. A recent paper [52] gave avery good review on this respect. It summarized the progress in forced-flow planarchromatography (FFPC) and demonstrated the importance of the differenttechniques like rotation planar chromatography (RPC), overpressured-layerchromatography (OPLC), and electroplanar chromatography (EPC). A simple, butpowerful preparative forced-flow technique was also reported; in this techniquehydrostatic pressure is used to increase mobile-phase velocity. Parallel- and serially-coupled layers open up new vistas for the analysis of a large number of samples (upto 216) for high throughput screening and for the analysis of very complex matrices.Some applications, relating to different classes of substances, were given todemonstrate the versatility of the various FFPC techniques [52].

2.2. Gas chromatography and volatile components inherbal medicinesIt is well-known that many pharmacologically active components in herbal medicinesare volatile chemical com56 Y.-Z. Liang et al. / J. Chromatogr. B 812 (2004) 5370pounds. Thus, the analysis of volatile compounds by gas chromatography is veryimportant in the analysis of herbal medicines. The GC analysis of the volatile oils hasa number of advantages. Firstly, the GC of the volatile oil gives a reasonablefingerprint which can be used to identify the plant. The composition and relativeconcentration of the organic compounds in the volatile oil are characteristic of theparticular plant and the presence of impurities in the volatile oil can be readilydetected. Secondly, the extraction of the volatile oil is relatively straightforward and

can be standardized and the components can be readily identified using GCMSanalysis. The relative quantities of the components can be used to monitor or assess


4/10

4

certain characteristics of the herbal medicines. Changes in composition of thevolatile oil may also be used as indicators of oxidation, enzymatic changes ormicrobial fermentation.The advantages of GC clearly lie in its high sensitivity of detection for almost all thevolatile chemical compounds. This is especially true for the usual FID detection

and GCMS. Furthermore, the high selectivity of capillary columns enablesseparation of many volatile compounds simultaneously within comparatively shorttimes. Thus, over the past decades, GC is a popular and useful analytical toolin the research field of herbal medicines [5362]. Especially, with the use ofhyphenated GCMS instrument, reliable information on the identity of thecompounds is available as well (see next section for more detail). However, the mostserious disadvantage of GC is that it is not convenient for its analysis of the samplesof polar and non-volatile compounds. For this, it is necessary to use tedious samplework-up which may include derivatization. Therefore, the liquid chromatographybecomes an another necessary tool for us to apply the comprehensive analysis ofthe herbal medicines.

2.3. High-performance liquid chromatographyHPLC is a popular method for the analysis of herbal medicines because it is easy tolearn and use and is not limited by the volatility or stability of the sample compound.In general, HPLC can be used to analyze almost all the compounds in the herbalmedicines. Thus, over the past decades, HPLC has received the most extensiveapplication in the analysis of herbal medicines [6376]. Reversed-phase (RP)columns may be the most popular columns used in the analytical separation ofherbal medicines. It is necessary to notice that the optimal separation condition forthe HPLC involves many factors, such as the different compositions of the mobilephases, their pH adjustment, pump pressures, etc. Thus, a good experimentaldesign for the optimal separation seems in general necessary [73,76].In order to obtain better separation, some new techniques have been recentlydeveloped in research field of liquid chromatography. These are micellarelectrokinetic capillary chromatography (MECC) [77], high-speed counter-currentchromatography (HSCCC), low-pressure size-exclusion chromatography (SEC) [78],reversed-phase ion-pairing HPLC (RP-IPC-HPLC) [79,80], and strong anion-exchange HPLC (SAX-HPLC) [81]. They will provide new opportunities for goodseparation for some specific extracts of some herbal medicines. On the other hand,the advantages of HPLC lie in its versatility for the analysis of the chemicalcompounds in herbal medicines, However, the commonly used detector in HPLC,say single wavelength UV detector,seems to be unable to fulfill the task, since lots of

chemical compounds in herbal medicines are non-chromophoric compounds.Consequently, a marked increase in the use of HPLC analysis coupled withevaporative light scattering detection (ELSD) in a recent decade demonstrated thatELSD is an excellent detection method for the analysis of non-chromophoriccompounds [8284]. This new detector provides a possibility for the direct HPLCanalysis of many pharmacologically active components in herbal medicines, sincethe response of ELSD depends only on the size, shape, and number of eluateparticles rather than the analysis structure and/or chromophore of analytes as UVdetector does. Especially, this technique is quite suitable for the constructionof the fingerprints of the herbal medicines. Moreover, the qualitative analysis orstructure elucidation of the chemical components in HM by simple HPLC is not

possible, as they rely on the application of techniques using hyphenated HPLC, such


5/10

5

as HPLCMS, HPLC-NMR, for the analysis of herbal medicines. This topic will befurther discussed later on.2.4. Electrophoretic methodsCapillary electrophoresis was introduced in early 1980s as a powerful analytical andseparation technique [85] and has since been developed almost explosively. It

allows an efficient way to document the purity/complexity of a sample and canhandle virtually every kind of charged sample components ranging from simpleinorganic ions to DNA. Thus, there was an obvious increase of electrophoreticmethods, especially capillary electrophoresis, used in the analysis of herbalmedicines [86108] in last decades. The more or less explosive development ofcapillary electrophoresis since its introduction has to a great extent paralleled that ofliquid chromatography. Most of the used techniques are capillary zoneelectrophoresis (CZE), capillary gel electrophoresis (CGE) and capillary isoelectricfocusing (cIEF). CE is promising for the separation and analysis of active ingredientsin herbal medicines, since it needs only small amounts of standards and can analyzesamples rapidly with a very good separation ability. Also, it is a good tool for

producing the chemical fingerprints of the herbal medicines, since it has similartechnical characteristics of liquid chromatography. Recently, several studies dealingwith herbal medicines, have been reported and two kinds of medicinal compounds,i.e. alkaloids [95102] and flavonoids [103108], have been studied extensively.In general, CE is a versatile and powerful separation tool with a high separationefficiency and selectivity when analyzing mixtures of low-molecular-masscomponents.Y.-Z. Liang et al. / J. Chromatogr. B 812 (2004) 537057However, as Shibabi and Hinsdale [109] pointed out that the fast development incapillary electrophoresis has so far been focused on the improvement of resolutionand throughput rather than reproducibility and absolute precision. One successfulapproach to improve the reproducibility of both mobility and integral data has beenbased on internal standards [110]. Many papers were published on the subject overthe last years but unfortunately they oftenonly give a limited image on the realpossibilities of CE in the field of fingerprinting herbal medicines. On the stony road toconsistent results, not only the complexity heterogeneity and polydispersity of herbalmedicines may limit straightforward conclusions, but also many artifacts can resultfrom the chosen separation buffer chemistrywith hidden instrumental constraints.However, we believe that CE and capillary electrochromatography approacheswould contribute to a better understanding of the solution behavior of herbalmedicines, especially when additionally combined (offline or online) with the powerful

spectrometric detectors.3. Hyphenation proceduresIn the past two decades, combining a chromatographic separation system on-linewith a spectroscopic detector in order to obtain structural information on the analytespresent in a sample has become the most important approach for the identificationand/or confirmation of the identity of target and unknown chemical compounds. Formost (trace-level) analytical problems in the research field of herbal medicines, thecombination of column liquid chromatography or capillary gas chromatography with aUVvis or a mass spectrometer (HPLCDAD, CE-DAD, GCMS and LCMS,respectively) becomes the preferred approach for the analysis of herbal medicines. Itis also true that additional and/or complementary information is, in quite a number of

cases, urgently required. This can be provided by, for example, atomic emission,Fourier-transform infrared (FTIR), fluorescence emission


6/10

6

(FE), or nuclear magnetic resonance (NMR) spectrometry. It is demonstrated that,from a practical point of view, rewarding results can be obtained, since we needmuch more information to deal with the most complex analytical systems such asthose samples from herbal medicines. Furthermore, the data obtained from suchhyphenated instruments are the so-called two-way data, say one way for

chromatogram and the other way for spectrum, which could provide much moreinformation than the classic one-way chromatography. With the help ofchemometrics, a rather new discipline developed both in chemistry and statistics inthe later part of the 1970s, we will definitely get more chance to deal with the difficultproblems in the analysis of herbal medicines and also the problems in quality controlof herbal medicines, which will be discussed in some detail in this review.3.1. GCMS and herbal medicinesMass spectrometry is the most sensitive and selective method for molecular analysisand can yield information on the molecular weight as well as the structure of themolecule. Combining chromatography with mass spectrometry provides theadvantage of both chromatography as a separation method and mass spectrometry

as an identification method. In mass spectrometry, there is a range of methods toionize compounds and then separate the ions. Common methods of ionization usedin conjunction with gas chromatography are electron impact (EI) and electroncapture ionization (ECI). EI is primarily configured to select positive ions, whereasECI is usually configured for negative ions (ECNI). EI is particularly useful for routineanalysis and provides reproducible mass spectra with structural information whichallows library searching. GCMS was the first successful online combination ofchromatography with mass spectrometry, and is widely used in the analysis ofessential oil in herbal medicines [111125].With the GCMS, people could produce not only a chromatographic fingerprint of theessential oil of the herbal medicine but also the information related to its mostqualitative and relative quantitative composition [116120,126,127]. Used in theanalysis of the herbal medicines, there are at least two significant advantages forGCMS, that is: (1) with the capillary column, GCMS has in general very goodseparation ability, which can produce a chemical fingerprint of high quality; (2) withthe coupled mass spectroscopy and the corresponding mass spectral database, thequalitative and relatively quantitative composition information of the herb investigatedcould be provided by GCMS, which will be extremely useful for the further researchfor elucidating the relationship between chemical constituents in herbal medicine andits pharmacology in further research. Thus, in our opinion, GCMS should be themost preferable tool for the analysis of the volatile chemicalcompounds in herbal

medicines.3.2. HPLCDAD, HPLCMS and othersHPLCDAD has become a common technique in most analytical laboratories in theworld now. With the additional UV spectral information, the qualitative analysis ofcomplex samples in herbal medicines turns out to be much easier than before. Forinstance, checking peak purity and comparing with the available standard spectrumof the known compound to the one in the investigated sample. Especially, with theintroduction of electrospray mass spectrometry, the coupling of liquidchromatography and mass spectrometry has opened the new way to widely androutinely applied to the analysis of herbal medicines. HPLC chromatographicfingerprints can be then applied for documentation of complete herbal extracts with

more information and on-line qualitative analysis becomes possible. Severalvaluable review articles dealing


7/10

7

Evaluation of chemical fingerprints of herbal medicines

In the early chemometric research, chromatographic data were commonly firsttransformed to retention timepeak area data matrices including only selectedpeaks, whether the identity of the peaks are known or not. The data such obtainedwere then used to do the processing, that is, the calculation of similarity or

dissimilarity between the fingerprints and the analysis of principal componentanalysis [184186].However, as pointed by Nielsen [176], The quality of the data (including onlyretention timepeak area data) relies on peak detection (integration) and on howthepeaks are selected for the data analysis. It can be very difficult to select an optimalset integration parameters for chromatograms obtained from analysis of complexsamples which easily contain more than 100 peaks. Furthermore, the selection andextraction of peaks to include in data analysis is difficult, partly subjective and a largeamount of the data in the chromatograms are discarded. The disadvantages of peakdetection and integration, and of the introduction of a subjective peak selection canbe avoided by using all collected data points in the chemometric analysis. Thus,

following the suggestion of Nielsen, the entire chromatographic profiles were utilizedto perform direct chemometric analysis. The analysis can be easily done with thehelp of proper techniques of data compression, such as the technique of wavelet orFourier transformation [187189], if necessary. Furthermore, another advantage oftaking the entire chromatographic profile to perform direct chemometric analysis isthat the peak shape can be included in data analysis, which will make thepretreatment of overlapping peaks much easier when one does evaluation of thefingerprints. Of course, the chromatographic profiles should properly alignedto compensate for minor drifts in retention times before one does the fingerprintevaluation and chemoemtric analysis for the purpose of quality control.5.1. Fingerprints and quality control in herbal medicinesAs stated above, one or two markers or pharmacologically active components werecurrently employed for evaluating the quality and authenticity of a herbal medicine.This kind of the determination, however, does not give a complete picture of anherbal product and therefore it will definitely fail to do the identification of false andtrue plant extraction. In the following an example will be given to illustrate thesituation.Fig. 3 shows 18 fingerprints of Ginkgo bilobaextractions,which were purchased from several pharmaceutical stores, vendors/companies andcollected from various producing areas in the mainland of PR China. All of thesesamples were supposed to meet the standard measured by UV spectroscopy at

wavelength of 318 nm with satisfactory absorbance (old standard method for qualitycontrol of G. Bilobaextraction in China), among which standard extract EGb761was friendly donated to one of authors from Guangzhou Institute for Drug Control,PR China by a Frenchman from Beaufour-Ipsen Company in France with asatisfactory fingerprint pattern. Analytical grade methanol and phosphoric acid usefor mobile phase were purchased from chemical reagents factory of Guangzhou,Guangdong, PR China. The other involved reagents were also of analytical grade.Ultrapure water (18.2M) was obtained by means of a Milli-Q apparatus by MilliporeCorporation (France) and was used for mobile phase preparation. The mobile phasewas vacuum filtered through a filter of 0.45 _m pore size. From the plot, it is difficultto find some false one. But if we simply do the PCA upon the fingerprints, the results


8/10

8

are shown in Fig. 4. It can be easily seen from the plot that samples marked bynumbers 13 are clearly outliers. Thus, if we pick up the fingerprints of samples 2and 3 (see Fig. 5C and D) and comparing them with the fingerprints (see Fig. 5Aand B) of the standard extract EGb761 (number 17 in Fig. 4) and the other sample(number 8 in Fig. 4), we can easily find the difference between them. The peak in the

fingerprints of samples 2 and 3 around the retention time of 10 min is much higherthan the one in the standard extract EGb761 and sample 8. This peak is rutin. Infact, rutin was added in the three outlier samples, say samples 13, in order to meetthe old standard of enough absorbance. They are quite different from the real G.biloba extractions as shown in Fig. 5A. From this example, we can see that thetechnique of fingerprint could really identify the false herbal products.5.2. Similarity of fingerprintsThe construction of chromatographic fingerprints aims at evaluating the quality ofHMs. As discussed above, the fundamental reason of quality control of herbalmedicines is based

Chemical pattern recognition and classification evaluationAs discussed above, the relationship within a set of chromatographic fingerprintscould be currently analyzed through comparison in terms of similarity or dissimilarityof the objects with a certain reference, presented as correlation coefficient orcongruence coefficient, etc. But, it has been aware that there are two problems forthis comparison: how to achieve the reasonable reference (comparing standard) andto what extent the investigated object is similar with the reference. Popularly, thereference may be derived either from standard extract of herbal medicine orproportioned mixture of herbal medicine (e.g. EGb761) or from computation by somemathematical methods (for the example above, the median chromatographicfingerprint of whole samples is taken, since there are three outliers in the samplesinvestigated). However, it is well known that natural products derived from herbalmedicines with inherent uncertainty feature of its 64 Y.-Z. Liang et al. / J.Chromatogr. B 812 (2004) 5370 secondary metabolic substances, to define anabsolute reference fingerprint by simply calculating their mean or median for onekind of herbal medicine seems somewhat subjective. From this point of view, theconception of class of one herbal medicine seems to be more reasonable. Thus, thechemical pattern recognition methods, such as K-nearest neighbors (KNN) [190,192]and soft independent modeling of class analogy (SIMCA) [193], etc. should be takeninto consideration for reasonable definition of the class of the herbal medicine [194196]. In fact, several researchers in China had worked on the concepts of using

chemical analytical and chromatographical fingerprinting to measure the consistencyof raw Chinese medicinal herbs and composite formula with the application of fuzzyclustering analysis of HPLC pattern in the early 1990s [225].On the other hand, the (dis)similarities of herbal objects with the reference oftenundertake themselves to a qualified threshold [191], which is not so easy to define.Although such a comparison attaches importance to the integral relationshipof the fingerprints, sometimes masking and swamping effects might occur eitherexplicitly or implicitly. The masking effect is that an unexpected sample is undecidedbecause of high similar value (e.g. the identification of three species of Coptischinensis, C. teet-Oides C. Y. Cheng, and C. deltoidea C. Y. Cheng et Hsiao fromherb Rhizoma coptidis). The swamping effect encompasses wrongly discriminating

a desirable sample illegal on account of low similarity with the reference influencedby the diversity of chromatographic compositional distribution (e.g. the determination


9/10

9

of herb Houttuynia cordataThunb. from different sources). To avoid these effects asmuch as possible, a method based on PCA after necessary data transformations[219]. The method has been demonstrated that PCA with standard normal variatetransformation of data led to meaningful classification of 33 different E. breviscapusherbal samples (see Fig. 2). The resultwas also corroborated by variance squares

discriminant method. The quality of herbal objects was further evaluated, and thecauses of this fact have been explained from a chemical point of view. The othermethod [220] is based on secured principal component regression (sPCR) that wasoriginally developed for detecting and correcting uncalibrated spectral features newlyemerging in spectra after the PCR calibration [196,197]. It can detect and considerunexpected chromatographic features for quality evaluation of herbal samples fromthe point of view of analyzing fingerprint residual.6. Qualityfication and validation of two-way datafrom hyphenated chromatographies by chemometricsIn general, the data generated by the hyphenated instruments are matrices withevery rowbeing a spectrum and every column a chromatogram at some wavelength,

wavenumber or m/e unit as illustrated in Fig. 6. The data obtained by suchhyphenated instruments in chemistry is generally called twoway or two-dimensionaldata. In common, the size of the data Fig. 6. Illustration of two-dimensional data fromhyphenated instruments. matrix such-obtained is rather big, sometimes it can bemore than 40 megabytes. Thus, this is a really a new challenge for modern analyticalchemistry to deal with the chemical information embedded in it.However, data derived from hyphenated analytical techniques have severaladvantages over the classic one-way chromatograms. Firstly, the two-way datamatrix contains both information from spectra and chromatograms, which will makeon-line structure identification of some interested compounds possible; secondly, thetwo-way data has so-called dimension advantages proposed by Booksh andKowalski [19], which will make on-line comparison of overlapping chromatographicpeaks possible; finally, the hyphenated technique might enhance thechromatographic separation ability by the additional spectral information, since onecould easily find some useful component selectivity with the help of chemometriclocal rank analysis methods, which can only be used for two-way chromatographicdata but never for one-way chromatographic data. For more detail, the readers canconsult reference [218]. How to use these advantages from the two-way data toaddress the problems in evaluation of the fingerprints of herbal medicines and theproblems in quality control of herbal medicines, which is just the topic we want todiscuss in the following subsections. There are, to our best knowledge, still

many different kinds of difficult problems unsolved in the research field of herbalmedicines. For instance, there may be no big problem for evaluating the fingerprintsfrom the same instruments and/or with the same batch of the herbal productsas shownabove in this review.However, ifweget some fingerprints from differentlaboratories and/or from the same kind of column, say C-18 for example, but fromdifferent companies. Could we still evaluate them reasonably? Moreover, could wedo the on-line comparison among the fingerprints of\ some overlappingchromatographic peaks and/or of big diversity? For instance, if we get somefingerprints from different extraction methods and/or from different herbal medicines,is it possible for us to see whether we have got the same phytochemical constituentsor not in order to understand their bioactivities and possible side effects of these

herbal products and consequently to enhance product quality control?With the help of the two-way data from the hyphenated chromatographies


10/10

10

and chemometric methods [198] recently developed,the answer is positive.6.1. Spectral correlative chromatogram and itsapplicationsAs stated above, it is very important for assessment of the quality of various samples

to determine the presence or absence of interested components among the differentchromatographic fingerprints. They may be obtained either from same herbs or fromdifferent ones under the same or similar chromatographic separated conditions.Moreover, another problem is the shift of retention time of some interested peaksof various fingerprints due to inevitable possibility in quality control, such as thefingerprints from different laboratories and the experimental columns from differentvendors despite the same type stationary-phase characteristics. This maybe lead toerroneous assess of quality of medical samples. Let us see an example shown inFig. 6.x1 and x2 shown in Fig. 7A are the chromatograms of twodimensionaldata sets ofX1 (of the G. bilobasamples obtained

Documents

The Traditional Herbal Medicines