AtW

La

b

a

ARAA

KPFDDH

1

iabfeptccuadteoan

0d

Journal of Analytical and Applied Pyrolysis 90 (2011) 13–17

Contents lists available at ScienceDirect

Journal of Analytical and Applied Pyrolysis

journa l homepage: www.e lsev ier .com/ locate / jaap

pplication of pyrolysis-gas chromatography and hierarchical cluster analysis tohe discrimination of the Chinese traditional medicine Dendrobium candidum

all. ex Lindl.

ili Wanga,∗, Cong Wanga,b, Zaifa Pana, Yang Suna, Xiangying Zhua

College of Chemical Engineering and Materials Science, Zhejiang University of Technology, Hangzhou, Zhejiang 310014, PR ChinaZhejiang Medicines & Health Products I/E Co., Ltd., Hangzhou, Zhejiang 310003, PR China

r t i c l e i n f o

rticle history:eceived 30 June 2008ccepted 30 September 2010vailable online 8 October 2010

eywords:

a b s t r a c t

The pyrogram fingerprints of Dendrobium candidum Wall. ex Lindl. samples from 3 different growingplaces and 2 other different species were analyzed on the basis of pyrolysis-gas chromatography. Anamount of 0.4 mg of sample powder was pyrolyzed in a vertical microfurnace pyrolyzer at 450 ◦C, and theproducts were directly introduced into a gas chromatograph equipped with a flame ionization detectoror a mass spectrometer. Then, each sample was characterized by the relative peak area of 40 peaks in

yrolysis-gas chromatographyingerprintendrobium candidum Wall. ex Lindl.endrobiumierarchical cluster analysis

thus obtained pyrogram. The pyrogram fingerprints of 16 samples from different growing places andspecies showed good similarity and reproducibility with the relative standard deviations (RSDs) of theretention time less than 0.12% (n = 5) and the RSDs of the relative percent of peak areas less than 3.77%(n = 5). Furthermore, the discrimination of the samples from different growing places and species wasachieved by hierarchical cluster analysis via recognizing the 16 × 40 data matrix. Thus, the results provedthe Py-GC fingerprint combined with chemometric approach is a simple, rapid and selective methodwhich is suitable for the quality control of the raw materials of herbal medicine.

. Introduction

Herbal medicines (HM) have been used for thousands of yearsn China and other countries because of their pharmacologicalctivities and low toxicity [1]. Dendrobium, known as ‘ShiHu’, haseen used in Chinese medicine therapy because it possesses theunctions of clearing heat, eyes-benefiting and immunomodulatoryffects [2]. There are 74 species and 2 variations of Dendrobiumlants found in China and about 30 species of them are used asraditional or folk medicines in China [3]. Among them, the mostommonly used and the most expensive species is Dendrobiumandidum Wall. ex Lindl. because of its excellent immunomod-latory effect. The chemical composition of Dendrobium whichffects its pharmacological activities often varies to some extentepending on the species and the growing places. However, it isroublesome and difficult to discriminate the D. candidum Wall.

x Lindl. materials from different species and identify the fakesf them by traditional methods such as microscopy by a botanistnd/or quantitative analysis of several main chemical compo-ents with HPLC or TLC. It is likely that the difference of multiple

∗ Corresponding author. Tel.: +86 571 88320416; fax: +86 571 88320103.E-mail addresses: lili wang@zjut.edu.cn, lili wang89@hotmail.com (L. Wang).

165-2370/$ – see front matter © 2010 Elsevier B.V. All rights reserved.oi:10.1016/j.jaap.2010.09.010

© 2010 Elsevier B.V. All rights reserved.

chemical components among sample is not significant enough fortheir discrimination by the above traditional methods. Therefore,a convenient, precise and holistic approach for quality control isnecessary.

Fingerprinting method using chromatographic technique forauthentication and quality control of Chinese Medicine Materi-als has recently been accepted by the World Health Organization(WHO) as a strategy for the assessment of herbal medicines [4]. Fur-thermore, chromatographic methods are highly recommended fordetermining the fingerprints of herbal medicine products and theirraw materials by the Drug Administration Bureau of China [5]. Sev-eral chromatographic fingerprint methods have been developed asuseful tools in the authentication and quality assessment of herbalmedicines [6–11], and HPLC fingerprinting analysis of the Dendro-bium species has also been reported [12–14]. However, prior to thefinal HPLC measurement pretreatments such as solvent extractionand concentration were inevitably utilized, which provided incom-plete components for a fingerprint analysis. In a word, a method forquality control requires the properties of speediness and automa-

tism, and the ability to analyze large amount of samples routinely.

Recently, pyrolysis-gas chromatography (/mass spectrometry)(Py-GC(/MS)) were used extensively in the characterization of nat-ural products such as lignin [15,16], food product [17] and botanicalmedicines [18,19]. This technique yields a pyrogram which consists

14 L. Wang et al. / Journal of Analytical and Applied Pyrolysis 90 (2011) 13–17

Table 1A summary of the test samples.

No. Sample code Sample species Source Time (year, month)

1 ZY-1 Dendrobium candidum Wall. ex Lindl. Yiwu, Zhejiang, China 2007, 12 ZY-2 Dendrobium candidum Wall. ex Lindl. Yiwu, Zhejiang, China 2007, 13 ZY-3 Dendrobium candidum Wall. ex Lindl. Yiwu, Zhejiang, China 2007, 14 ZY-4 Dendrobium candidum Wall. ex Lindl. Yiwu, Zhejiang, China 2007, 15 ZT-1 Dendrobium candidum Wall. ex Lindl. Tiantaishan, Zhejiang, China 2007, 16 ZT-2 Dendrobium candidum Wall. ex Lindl. Tiantaishan, Zhejiang, China 2007, 17 ZT-3 Dendrobium candidum Wall. ex Lindl. Tiantaishan, Zhejiang, China 2007, 18 AH-1 Dendrobium candidum Wall. ex Lindl. Huoshan, Anhui, China 2007, 39 AH-2 Dendrobium candidum Wall. ex Lindl. Huoshan, Anhui, China 2007, 3

10 AH-3 Dendrobium candidum Wall. ex Lindl. Huoshan, Anhui, China 2007, 311 GG-1 Dendrobium crystallinum Rchb. f. Guangzhou, Guangdong, China 2007, 312 GG-2 Dendrobium crystallinum Rchb. f. Guangzhou, Guangdong, China 2007, 4

owodc

mgtgssps

2

2

sacdi

imt

Fa

of the pyrolyzer kept at around room temperature, and then it wasdropped into the heated center of the pyrolyzer under the flowof nitrogen carrier gas. In order to obtain good peak intensity andappropriate peak number, the optimum pyrolysis temperature of

1 (ZY-1)

2 (ZY-2)

3 (ZY-3)

4 (ZY-4)

5 (ZT-1)

6 (ZT-2)

7 (ZT-3)

13 GG-3 Dendrobium crystallinum Rchb. f.14 YS-1 Dendrobium devonianum Paxt.15 YS-2 Dendrobium devonianum Paxt.16 YS-3 Dendrobium devonianum Paxt.

f the characteristic peaks of the constituents in a given sampleithout any pretreatment procedures. Based on the peak intensity

bserved on the pyrogram, not only chemical composition but alsoiscriminative analysis is achieved by further combination withhemometric approaches.

In this study, pyrolysis-gas chromatography using a verticalicrofurnace pyrolyzer was applied for construction of Py-GC fin-

erprint for D. candidum Wall. ex Lindl. raw material without anyedious pretreatments. Furthermore, on the basis of Py-GC fin-erprint discriminative analyses for D. candidum Wall. ex Lindl.amples from 3 different growing places and 2 other differentpecies were performed by chemometric approaches on the basis ofrincipal component analysis (PCA) and hierarchical cluster analy-is (HCA).

. Experimental

.1. Material

Total 16 samples were used in this study. As shown in Table 1, 10amples were D. candidum Wall. ex Lindl. collected from Zhejiangnd Anhui province, 3 samples were Dendrobium devonianum Paxt.ollected from Yunnan province, and other 3 samples were Den-robium crystallinum Rchb. f. collected from Guangdong provincen China.

All samples were dried at 60 ◦C for 6 h, and then were millednto fine powders (finer than 120 meshes) by an herbal medicine

ill (HX-100) prior to Py-GC measurements in order to improvehe efficiency of the pyrolysis.

1

7

11

1618

20

21

25

26

27

2830

29

3132

3334

35

36

37

38

39

605040302010

2

3

4

5

6

8

109

12

13

1415

17

19

22

23

24

t / min0 65

40

ig. 1. A typical pyrogram of Dendrobium candidum Wall. ex Lindl. sample obtainedt pyrolytic temperature of 450 ◦C.

Guangzhou, Guangdong, China 2007, 5Simao, Yunnan, China 2007, 5Simao, Yunnan, China 2007, 6Simao, Yunnan, China 2007, 7

2.2. Py-GC condition

A vertical microfurnace pyrolyzer (PY2020iD, Frontier Lab Ltd.,Fukushima, Japan) was directly attached to a gas chromatograph(CP-3800, Varian, USA) equipped with a flame ionization detec-tor (FID). About 0.4 mg of powdered Dendrobium sample taken ina platinum sample cup was first mounted at the waiting position

8 (AH-1)

9 (AH-2)

10 (AH-3)

11 (GG-1)

12 (GG-2)

13 (GG-3)

14 (YS-1)

15 (YS-2)

16 (YS-3)

605040302010t/ min

0

Fig. 2. Typical pyrograms of 16 samples obtained at pyrolytic temperature of 450 ◦C.(Samples 1–10 are Dendrobium candidum Wall. ex Lindl., samples 11–13 are Den-drobium crystallinum Rchb. f., samples 14–16 are Dendrobium devonianum Paxt.)

L. Wang et al. / Journal of Analytical and Applied Pyrolysis 90 (2011) 13–17 15

Table 2Peak assignments of the main components for Dendrobium candidum Wall. ex Lindl.

Peak no. tR (min) RSD (%)(n = 5)for tR

Compound Molecularformula

Molecularweight

Match Peak area (%) RSD (%) forPA (n = 5)

1 2.29 0.07 Carbon dioxide CO2 44 – 7.91 0.632 2.78 0.09 Formic acid CH2O2 46 86 6.38 0.283 3.63 0.08 2,3-Butanedione C4H6O2 86 80 1.77 2.104 4.20 0.04 Acetic acid C2H4O2 60 90 8.12 2.195 4.59 0.07 Methyl formate C2H4O2 60 78 15.64 2.546 6.66 0.12 1-Hydroxy-2-propanone C3H6O2 74 72 10.89 0.737 10.20 0.09 Dihydro-2-methyl-3(2H)-furanone C5H8O2 100 86 0.69 3.538 11.33 0.06 2-Methyl-2-cyclopenten-1-one C6H8O 96 87 2.60 1.489 11.68 0.05 4(1H)-Pyridinone C5H5NO 95 88 0.66 1.34

10 11.88 0.06 2-Methyl-furan C5H6O 82 91 0.66 1.6911 12.47 0.07 Cyclohexanone C6H10O 98 87 1.98 1.2612 12.90 0.06 Phenol C6H6O 94 90 4.06 0.7213 13.43 0.03 Butyrolactone C4H6O2 86 86 1.42 0.0414 13.59 0.07 3-Methyl-2-cyclopenten-1-one C6H8O 96 93 1.28 3.2715 13.85 0.09 Furfural C5H4O2 96 96 1.04 2.6916 15.06 0.06 o-Methylisourea C2H6N2O 74 78 0.55 2.8917 15.19 0.08 1-(Dimethylamino)pyrrole C6H10N2 110 85 0.55 1.518 15.91 0.02 1,2-Dimethyl-cyclohexene C8H14 110 98 0.80 2.6719 16.96 0.10 2-Hydroxy-3-methyl-2-cyclopenten-1-one C6H8O2 112 70 2.67 2.7120 17.41 0.05 4-Methyl-phenol C7H8O 108 98 1.08 2.7521 18.23 0.06 Dihydro-2,4(1H,3H)-pyrimidinedione C4H6N2O2 114 80 0.83 1.4322 18.45 0.04 3-Methyl-1,2-cyclopentanedione C6H8O2 112 93 1.08 1.8623 18.74 0.04 Mequinol C7H8O2 124 94 1.98 2.1524 20.07 0.06 3-Ethyl-2-hydroxy-2-cyclopenten-1-one C7H10O2 126 70 2.12 3.3225 21.49 0.04 2,5-Dimethyl-cyclopentanone C7H12O 112 91 1.63 3.3726 24.28 0.03 l-Guanidinosuccinimide C5H7N3O2 141 80 3.64 2.9927 25.74 0.05 2,3-Dihydro-benzofuran C8H8O 120 92 0.59 3.2728 26.42 0.03 4-Ethyl-2-methoxy-phenol C9H12O2 152 96 0.55 3.6829 26.66 0.03 Hexahydro-1-nitroso-1H-azepine C6H12N2O 128 86 0.80 1.7230 27.27 0.04 3-Methyl-2-hexene C7H14 98 90 0.66 2.4231 27.97 0.03 Vanillin C8H8O3 152 95 3.26 2.1732 30.10 0.02 2,6-Dimethoxy-phenol C8H10O3 154 94 1.87 3.3833 31.35 0.02 2-Methoxy-4-(1-propenyl)-phenol C10H12O2 164 94 0.73 3.1434 32.10 0.02 3-Hydroxy-4-methoxybenzoic acid C8H8O4 168 90 0.73 1.4735 35.13 0.02 4-Methyl-2,5-dimethoxybenzaldehyde C10H12O3 180 86 2.60 1.7536 35.60 0.02 1,6-Anhydro-.beta.-d-glucopyranose (levoglucosan) C6H10O5 162 90 1.42 0.47

dr9-one

4pw3urtttaoatir9maif

2a

b

37 38.02 0.02 2,6-Dimethoxy-4-(2-propenyl)-phenol38 39.86 0.03 3,5-Dimethoxy-4-hydroxyphenylacetic aci39 45.47 0.03 N-(2-hydroxybenzoyl)-glycine methyl este40 48.53 0.02 4,5-Dihydroxy-2,7-dimethoxy-9H-fluoren-

50 ◦C was empirically determined after examining various tem-eratures between 300 and 600 ◦C. A silica capillary column coatedith 50% diphenyl–50% dimethylpolysiloxane (Varian CP-SIL 24 CB,

0 m × 0.25 mm i.d., 0.25 �m film thickness) in its interior wall wassed. The flow rate of 30 ml/min of carrier gas at the pyrolyzer waseduced to 1.0 ml/min at the capillary column by means of a split-er. The column temperature was initially set at 35 ◦C (keep 5 min),hen heated from 35 ◦C to 260 ◦C at a rate of 5 ◦C/min. The finalemperature was held for 15 min. Both the injector temperaturend the detector temperature were kept at 300 ◦C. Identificationf the peaks on the resulting pyrograms was carried out by use ofGC/MS system (HP 5890/5972, Hewlett-Packard, USA) to which

he same pyrolyzer was also attached. For the MS measurement,onization was executed by electron impact (EI) at 70 eV. The massange m/z 40–500 was scanned at a frequency of 2 scan s−1. The NIST8 and WILEY 138 spectra libraries were used. The peak assign-ents were carried out through direct comparison of commercially

vailable mass spectra database using the dedicated library search-ng system together with the interpretation of their mass spectralragmentation patterns.

.3. Principle component analysis (PCA) and hierarchical clusternalysis (HCA)

The resulting data from the observed pyrograms was processedy normalization of the selected peak areas, and then was subjected

C11H14O3 194 94 1.70 3.33C10H12O5 212 87 1.42 2.22C10H11NO4 209 85 0.87 3.77C15H12O6 272 90 0.76 2.69

to analysis with principle component analysis (PCA) and hierarchi-cal cluster analysis (HCA). The PCA and HCA of samples 1–16 wereperformed using SPSS software (SPSS for Windows 11.5, SPSS Inc.,USA). A method called average linkage between groups was appliedand Pearson correlation was selected as a measurement in HCA.

3. Results and discussion

3.1. Pyrogram of D. candidum Wall. ex Lindl. and methodvalidation

Fig. 1 shows a typical pyrogram of D. candidum Wall. ex Lindl.sample obtained at an optimum pyrolytic temperature of 450 ◦C. Onthe pyrogram of Fig. 1, a series of sharp peaks of pyrolyzates with farbetter resolution were observed. By similarity match using the massspectrometric libraries, a majority of the chemical componentswere identified. The assigned peaks on this pyrogram are listedin Table 2 together with their retention time (tR), relative stan-dard deviation (RSDs) for tR, molecular formula, molecular weight,the match quality between the experimental spectrum and that ofthe database used for the identification, relative percent of peak

areas (RPPA) and RSDs for RPPA. Here, the main components atretention time from 10 to 40 min were identified as aromatics suchas bibenzyls, phenanthrenes, fluorenones and coumarins formedpartly from the components of cellulose and lignin, and partlyfrom the pyrolyzates of pharmacological activities such as dehydra-

16 L. Wang et al. / Journal of Analytical and Applied Pyrolysis 90 (2011) 13–17

Table 3Relative peak area of 40 components for 16 samples.

Peak no. tR (min) Relative peak areaa

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16ZY-1 ZY-2 ZY-3 ZY-4 ZT-1 ZT-2 ZT-3 AH-1 AH-2 AH-3 GG-1 GG-2 GG-3 YS-1 YS-2 YS-3

1 2.29 2.29 2.02 1.98 2.04 1.55 1.56 1.44 1.80 1.85 1.75 1.19 1.16 1.20 2.05 2.08 1.982 2.78 1.84 1.85 1.70 1.76 1.35 1.36 1.34 1.39 1.44 1.36 0.80 0.81 0.79 1.63 1.63 1.583 3.63 0.51 0.45 0.45 0.46 0.36 0.37 0.38 0.48 0.51 0.49 0.24 0.24 0.24 0.26 0.27 0.264 4.20 2.34 2.02 2.12 2.11 1.43 1.44 1.36 1.80 1.86 1.82 2.99 2.95 2.95 1.45 1.51 1.485 4.59 4.52 3.95 3.69 3.96 3.04 2.80 2.78 2.41 2.50 2.33 2.60 2.49 2.55 3.80 3.78 3.916 6.66 3.14 2.86 2.89 2.91 2.06 2.16 1.98 2.25 2.30 2.25 2.16 2.13 2.14 1.90 1.92 1.857 10.20 0.20 0.19 0.20 0.20 0.09 0.10 0.10 0.09 0.09 0.09 0.14 0.14 0.14 0.16 0.16 0.168 11.33 0.75 0.71 0.68 0.70 0.48 0.46 0.50 0.60 0.62 0.59 0.46 0.44 0.46 0.57 0.58 0.569 11.68 0.19 0.17 0.17 0.17 0.12 0.11 0.12 0.12 0.12 0.12 0.11 0.11 0.11 0.20 0.19 0.18

10 11.88 0.19 0.16 0.17 0.17 0.12 0.10 0.12 0.15 0.16 0.15 0.08 0.08 0.08 0.23 0.27 0.2611 12.47 0.57 0.50 0.50 0.51 0.35 0.34 0.33 0.49 0.50 0.47 0.33 0.33 0.33 0.63 0.67 0.6512 12.90 1.17 1.04 1.02 1.05 0.78 0.79 0.79 0.85 0.88 0.85 0.90 0.90 0.87 0.48 0.49 0.4913 13.43 0.41 0.37 0.35 0.37 0.23 0.23 0.21 0.23 0.23 0.23 0.25 0.24 0.24 0.39 0.40 0.3914 13.59 0.37 0.34 0.35 0.34 0.46 0.46 0.45 0.43 0.43 0.43 0.59 0.58 0.59 0.01 0.01 0.0115 13.85 0.30 0.26 0.26 0.27 0.24 0.25 0.23 0.25 0.26 0.26 0.27 0.26 0.28 0.16 0.17 0.1616 15.06 0.16 0.15 0.15 0.15 0.12 0.13 0.13 0.13 0.13 0.13 0.15 0.14 0.14 0.18 0.19 0.1817 15.19 0.16 0.15 0.14 0.15 0.10 0.11 0.10 0.10 0.10 0.10 0.12 0.12 0.12 0.15 0.17 0.1618 15.91 0.23 0.20 0.19 0.20 0.08 0.09 0.08 0.15 0.15 0.16 0.15 0.16 0.15 0.23 0.24 0.2219 16.96 0.77 0.74 0.73 0.73 0.73 0.73 0.73 0.76 0.79 0.80 0.82 0.83 0.86 0.70 0.70 0.6820 17.41 0.31 0.26 0.28 0.28 0.25 0.24 0.24 0.28 0.29 0.28 0.24 0.23 0.23 0.42 0.43 0.4121 18.23 0.24 0.19 0.21 0.21 0.14 0.14 0.13 0.17 0.16 0.16 0.06 0.06 0.06 0.33 0.35 0.3422 18.45 0.31 0.29 0.27 0.28 0.32 0.32 0.31 0.32 0.32 0.31 0.39 0.38 0.39 0.51 0.52 0.5223 18.74 0.57 0.52 0.50 0.52 0.42 0.38 0.41 0.45 0.49 0.48 0.46 0.45 0.46 0.09 0.09 0.0924 20.07 0.61 0.56 0.58 0.56 0.42 0.43 0.44 0.43 0.43 0.44 0.56 0.57 0.57 0.17 0.18 0.1625 21.49 0.47 0.45 0.43 0.44 0.29 0.28 0.27 0.23 0.25 0.24 0.23 0.23 0.23 0.34 0.33 0.3326 24.28 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.0027 25.74 0.17 0.16 0.16 0.16 0.10 0.10 0.10 0.00 0.00 0.00 0.07 0.07 0.07 0.37 0.37 0.3628 26.42 0.16 0.15 0.16 0.15 0.10 0.10 0.10 0.10 0.10 0.10 0.10 0.09 0.10 0.30 0.32 0.3029 26.66 0.23 0.20 0.20 0.21 0.15 0.15 0.16 0.18 0.18 0.17 0.15 0.15 0.15 0.25 0.26 0.2430 27.27 0.19 0.19 0.19 0.19 0.09 0.10 0.10 0.08 0.08 0.09 0.17 0.17 0.17 0.21 0.21 0.2131 27.97 0.94 0.85 0.83 0.85 0.55 0.53 0.52 0.57 0.56 0.55 0.42 0.39 0.41 0.80 0.85 0.8032 30.10 0.54 0.50 0.45 0.48 0.16 0.16 0.16 0.00 0.00 0.00 0.09 0.09 0.09 0.30 0.32 0.3033 31.35 0.21 0.20 0.20 0.20 0.09 0.10 0.09 0.00 0.00 0.00 0.06 0.06 0.06 0.17 0.18 0.1734 32.10 0.21 0.20 0.19 0.20 0.00 0.00 0.00 0.09 0.09 0.09 0.08 0.08 0.08 0.08 0.08 0.0735 35.13 0.75 0.68 0.66 0.68 0.22 0.24 0.20 0.10 0.10 0.10 0.12 0.11 0.12 0.30 0.31 0.3036 35.60 0.41 0.32 0.33 0.35 0.07 0.08 0.13 0.11 0.11 0.11 0.15 0.15 0.15 0.18 0.19 0.1837 38.02 0.49 0.48 0.44 0.45 0.16 0.17 0.15 0.10 0.10 0.10 0.07 0.07 0.08 0.20 0.20 0.1938 39.86 0.41 0.35 0.35 0.36 0.08 0.15 0.18 0.00 0.00 0.00 0.10 0.09 0.09 0.13 0.13 0.14

0708

ea of t

t[

omvr3wotdtfi

3

WaFaaia

39 45.47 0.25 0.23 0.23 0.24 0.09 0.08 0.40 48.53 0.22 0.24 0.23 0.25 0.10 0.08 0.

a The relative areas of the 40 characteristic peaks were calculated by using the ar

ions of glycosides, alkaloids, phenanthrenes and polysaccharides20–22].

In order to achieve a stable and reproducible Py-GC fingerprintf D. candidum Wall. ex Lindl. for the purpose of quality assessment,ethod reproducibility was evaluated firstly by analyzing five indi-

idual samples over the period of investigation. The RSDs for tR andelative percent of peak areas were found to be less than 0.12 and.77% (n = 5) as shown in Table 2. Then, stability test was carried outith a D. candidum Wall. ex Lindl. over a period of one month. The

btained RSDs of tR and relative percent of peak areas were foundo be less than 0.12 and 3.85% (n = 5), illustrating the good repro-ucibility of the method. All of the above results suggested thathe developed methodology is applicable for establishing a Py-GCngerprint of D. candidum Wall. ex Lindl.

.2. Py-GC fingerprint of Dendrobium sample

To establish a representative Py-GC fingerprint for D. candidumall. ex Lindl., 16 samples of Dendrobium with different species

nd growing places were analyzed using the above Py-GC method.

ig. 2 shows typical pyrograms of 16 Dendrobium samples obtainedt the same pyrolytic conditions. As shown in this figure, althoughll pyrogram patterns are similar to one another, some variationsn peak abundance exist. It seems that it is possible to visu-lly differentiate the pyrograms based on their variations in peak

0.00 0.00 0.00 0.07 0.06 0.06 0.10 0.09 0.100.11 0.12 0.11 0.08 0.09 0.08 0.15 0.15 0.15

he reference peak (No. 26) as a reference standard.

abundance; however, the process is subjective and not quantita-tive. Furthermore, minor differences among the similar pyrogramsmight be missed. Hence, principle component analysis and hierar-chical cluster analysis were used here to compare and discriminatethe Dendrobium samples based on their pyrograms.

3.3. Discriminative analysis of Dendrobium samples fromdifferent growing place and species

Among the 16 obtained pyrograms, peaks having matched tRwith a reasonable abundance of the relative percent of peak areawere chosen and assigned as common peaks for representing thecharacteristic pattern of Dendrobium. Altogether 40 characteris-tic peaks were selected in 16 pyrograms (see Fig. 1). The peakat retention time 24.28 min (No. 26) was selected as a referencepeak because it was a strong single peak in the middle part of thepyrograms of all the 16 samples. The relative areas of the 40 char-acteristic peaks were calculated by using the area of the referencepeak (No. 26) as a reference standard. Relative areas of the 40 peaksof samples 1–16 formed a matrix of 16 × 40 shown in Table 3.

At first, principle component analysis was applied based on thedatabase listed in Table 3. After converting them into the devia-tion values, the contributory rates (dispersion) of the first, second,and third principal components obtained were 56%, 29% and 9%,respectively. Fig. 3 shows the relationship among the first, second

L. Wang et al. / Journal of Analytical and

Fig. 3. Results of principle component analysis of 16 Dendrobium samples.

Fg

ahtsst

gica

[

[[[

[

[[

[[

[469.

ig. 4. Results of hierarchical cluster analysis of 16 Dendrobium samples (dendro-ram using average linkage between groups).

nd third principal component scores for 16 samples. As shownere, these samples are divided into different groups according toheir different growing places and species. However, the group ofample ZT is more close to the group of sample AH than that ofample ZY even though samples of ZY and ZT are collected fromhe same province.

Hierarchical clustering is the process of the subdivision of aroup of samples into clusters that exhibit a high degree of bothntracluster similarity and intercluster dissimilarity. The outputan be represented as a dendrogram. The between-groups link-ge method, with the Cosine as the similarity measure, was used

[

[

[

Applied Pyrolysis 90 (2011) 13–17 17

in this work. Similarities among the 16 Dendrobium samples werecalculated based on the data in Table 3 using the SPSS software.The obtained dendrogram is shown in Fig. 4. It was clear that thesamples could be divided into 3 groups: samples 1–10 in cluster a,samples 14–16 in cluster b and samples 11–13 in cluster c, reflect-ing the difference of their species, i.e., D. candidum Wall. ex Lindl., D.devonianum Paxt. and D. crystallinum Rchb. f., respectively. In addi-tion, in the cluster of D. candidum Wall. ex Lindl. (samples 1–10),samples 1–4 and samples 5–7 were clear grouped together due totheir same origin place, i.e., Zhejiang province in China.

4. Conclusion

Pyrolysis-gas chromatography has been proved to be a rapid,convenient and highly selective method for construction of Py-GCfingerprint for D. candidum Wall. ex Lindl. raw material withoutusing any tedious pretreatments. The results show that not only thedifferent species of Dendrobium but also the D. candidum Wall. exLindl. from different growing places can be discriminated by usingHCA carried out on the basis of the relative area of the characteristicpeaks on Py-GC fingerprint. This method combined with chemo-metric approach is proved to be a promising tool for assessment ofthe quality of the raw materials of herbal medicine.

Acknowledgement

Financial support from the Natural Science Foundation of Zhe-jiang Province (No. Y307602) is gratefully acknowledged.

References

[1] K.C. Wen, C.Y. Huang, F.L. Lu, J. Chromatogr. A 631 (1993) 241.[2] Jiangsu New Medical College, Dictionary of Chinese Traditional Medicines,

Jiangsu New Medical College, Shanghai, 1986.[3] G.N. Zhang, Z.M. Bi, Z.T. Wang, L.S. Xu, G.J. Xu, Chin. Tradit. Herb. Drugs 34

(2003) S5.[4] World Health Organization, Guidelines for the Assessment of Herbal Medicines,

Munich, 28.6.1991, WHO, Geneva, 1991.[5] The State Drug Administration of China, Technical Requirements for Studying

Fingerprint of Traditional Chinese Medicine Injection (Draft), 2000.[6] A.J. Lau, S.O. Woo, H.L. Koh, J. Chromatogr. A 1011 (2003) 77.[7] A.J. Lau, B.H. Seo, S.O. Woo, H.L. Koh, J. Chromatogr. A 1057 (2004) 141.[8] F. Gong, Y.Z. Liang, P.S. Xie, F.T. Chau, J. Chromatogr. A 1002 (2003) 25.[9] Y.N. Ni, Y.Y. Peng, S. Kokot, Anal. Chim. Acta 616 (2008) 19.10] W. Li, Y.L. Deng, R.J. Dai, Y.H. Yu, M.K. Saeed, L. Li, W.W. Meng, X.S. Zhang, J.

Pharm. Biomed. Anal. 45 (2007) 38.11] Y.N. Ni, Y.Y. Peng, S. Kokot, Chromatographia 67 (2008) 211.12] Z.J. Zhang, Y.Y. Wang, Q. Li, J. Chin. Pharm. Univ. 34 (2003) 534.13] Z.J. Zhang, Y.Y. Wang, Q. Li, X.W. Wang, X.Y. Xu, H. Zhang, Chin. Tradit. Herb.

Drugs 35 (2004) 393.14] Y.Y. Wang, Z.J. Zhang, X.X. Li, X.W. Wang, H. Zhang, Chin. J. Pharm. Anal. 26

(2006) 568.15] N. Takeshi, H. Kawamoto, S. Saka, J. Anal. Appl. Pyrolysis 81 (2008) 173.16] J.C. del Río, A. Gutiérrez, I.M. Rodríguez, D. Ibarra, Á.T. Martínez, J. Anal. Appl.

Pyrolysis 79 (2007) 39.17] C. Pérès, C. Viallon, J.L. Berdagué, J. Anal. Appl. Pyrolysis 65 (2002) 161.18] Z.D. Zeng, Y.Z. Liang, T. Zhang, F.T. Chau, Y.L. Wang, Anal. Bioanal. Chem. 385

(2006) 392.19] M. Yuan, M.G. Zhang, P. Yuan, Z. Zeng, C.L. Yan, Chin. J. Chromatogr. 21 (2003)

20] G.N. Zhang, L.Y. Zhong, S.W. Annie Bligh, Y.L. Guo, C.F. Zhang, M. Zhang, Z.T.Wang, L.S. Xu, Phytochemistry 66 (2005) 1113.

21] L. Yang, L.H. Qin, S.W. Annie Bligh, A. Bashall, C.F. Zhang, M. Zhang, Z.T. Wanga,L.S. Xu, Bioorgan. Med. Chem. 14 (2006) 3496.

22] L. Yang, Z.T. Wang, L.S. Xu, Biochem. Syst. Ecol. 34 (2006) 658.