J. Sep. Sci. 2008, 31, 321 – 328 L. Wang et al.

321

Longhu Wang Short Communication

Yanting Song
Yiyu Cheng
Xuesong Liu
Department of Chinese Medicine
Sciences and Engineering,
College of Pharmaceutical
Sciences, Zhejiang University,
Hangzhou, China
Orthogonal array design for the optimization of
supercritical fluid extraction of tanshinones from
Danshen
Different from the west medicine, the therapeutic effect of the traditional Chinese
medicine is usually based on multifarious essential components or the combination
of them instead of only one component. In this paper, a novel supercritical fluid
extraction (SFE) method has been developed for extracting tanshinones (dihydrotan-
shinone I, cryptotanshinone, tanshinone I, and tanshinone IIA) from Danshen, the
dried root of Salvia miltiorrhiza. Various experimental conditions were investigated
to optimize the SFE. Under the appropriate conditions, extracting at 40 MPa, and
508C and with CO2 flow rate of 25 L/h for 1 h, SFE can achieve a better yield as well
as the recoveries of the tanshinones than the conventional extraction using metha-
nol. Moreover, the target compounds were analyzed by HPLC with a C18 RP column
by gradient elution using ACN and water as mobile phase at a flow rate of 1.0 mL/
min and with UV detection at 270 nm. Four calibration equations were then estab-
lished and good linear relationships were shown (r2 A0.999) in the investigated con-
centration range. The recoveries, measured at three concentration levels, varied
from 97.2 to 103.8%. The method provided in this article could be applied as an
improved quality control method for Danshen products.
Keywords: Salvia miltiorrhiza / Supercritical fluid extraction / Tanshinone / Traditional Chinese
medicine /
Received: August 12, 2007; revised: November 2, 2007; accepted: November 2, 2007
DOI 10.1002/jssc.200700386

1 Introduction Fig. 1) [6, 7]. Pharmacological experiments demonstrated

Danshen, the dried root of Salvia miltiorrhiza, has been
widely used either alone or in combination with other
herbal ingredients for the treatment of coronary artery
disease, hyperlipidemia, hypertension, arrhythmias, and
other clinical diseases in China and other countries
including the United States [1, 2]. According to the phar-
macological investigations, the active constituents of
Danshen could be classified as the lipid-soluble tanshi-
nones that belong to abietane-type diterpenes and the
water-soluble phenolic acids [3 – 5]. In recent years, many
active constituents were found in the lipid-soluble
extract of Danshen, such as dihydrotanshinone I, crypto-
tanshinone, tanshinone I, and tanshinone IIA (shown in
Correspondence: Dr. Xuesong Liu, Department of Chinese Medi-
cine Sciences and Engineering, College of Pharmaceutical Scien-
ces, Zhejiang University, Hangzhou 310027, China
E-mail: liuxuesong@zju.edu.cn
Fax: +86-571-88208426
Abbreviations: ANOVA, analysis of variance; OAD, orthogonal
array design; SE, solvent extraction; SFE, supercritical fluid ex-
traction; TCM, Traditional Chinese medicine
that dihydrotanshinone I possesses the function for the
arrest of the growth and apoptosis induction of cancer
cells, while cryptotanshinone is usually used to cure
against inflammation, tanshinone I is for therapy of
angina pectoris, and tanshinone IIA is for improving
blood circulation [2, 8].
Since conventional liquid – solid extraction tech-
niques, such as Soxhlet extraction, sonication, mechani-
cal shaking, and microwave-assisted extraction [9 – 11],
are tedious, time-consuming, and unfavorable for con-
sumption of hostile organic solvents and degradation of
thermally labile compounds, much attention has been
paid to developing new alternative techniques for the
rapid extraction from solid matrices, such as supercriti-
cal fluid extraction (SFE) [12, 13].
SFE has been emerging as a promising technique for
the extraction of the nonpolar compounds from herbs
used as traditional Chinese medicines (TCMs). Compar-
ing to the conventional extraction methods, SFE has
numerous advantages such as less consumption of hos-
tile organic solvents, reduction of the extraction time,
avoidance of degradation of active compounds, and pro-

i 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.jss-journal.com

322 L. Wang et al.
Figure 1. Molecular structures of four tanshinones.
duction of cleaner extracts [14]. The most frequently
employed supercritical solvent is CO2 due to its nonflam-
mable, nontoxic, and low critical temperature [15].
Because of the fat-soluble characteristic, tanshinones
are especially appropriate to be extracted by SFE. Dean et
al. [16] have studied the extraction of tanshinone IIA from
Danshen by SFE. However, they neglected the other coex-
tracted active compounds with similar effects. Different
from the western medicine, the therapeutic effect of the
TCM is usually based on multifarious essential compo-
nents or the combination of them instead of only using
one component. The evident example of Danshen prod-
ucts is Danshen Pill, which is widely used in China.
Therefore, it will be advantageous to take into account
multifold active chemical constituents while an SFE tech-
nique is developed. In this paper, the orthogonal test
design L9 (3)4 was designed to optimize the SFE of the four
commonly used tanshinones from the dried root of S. mil-
tiorrhiza. In addition, interactions among factors were
not incorporated in the matrix and the attention was
given upon the main effects of the four factors such as
extraction pressure and temperature, dynamic extrac-
tion time, and flow rate of CO2.
For evaluation of the SFE process, it is also necessary to
develop an HPLC method for simultaneously determin-
ing the investigated compounds in the product. Up to
now, large amounts of methods were established such as
HPLC-DAD-ESI MS [17, 18], high-speed countercurrent
chromatography (HSCCC) [19, 20], and HPLC-UV [21 – 23]
i 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
J. Sep. Sci. 2008, 31, 321 – 328
for the determination of tanshinones or other bioactive
constituents in Danshen. And some methods for simulta-
neously determining the four tanshinones in the prepa-
rations of TCM have been reported recently [24], but a rel-
atively higher standard of chromatography conditions
was required during the analysis. So, another aim of this
paper is to develop a simultaneous determination
method that can be used as a feasible analysis for
Danshen products.
2 Experimental
2.1 Materials and reagents
Dried root of S. miltiorrhiza was purchased from Tongren-
tang Medicine (Hangzhou, China) and cut into pieces
with diameter of about 3 mm and thickness of 3 – 5 mm.
HPLC-grade ACN and methanol were bought from
Merck (Darmstadt, Germany). Water for HPLC was puri-
fied by using a Milli-Q academic water purification sys-
tem (Milford, MA, USA). Reference standards of dihydro-
tanshinone I, cryptotanshinone, tanshinone I, and tan-
shinone IIA were purchased from Zhejiang Chemical
Reagent Plant (Hangzhou, China). CO2 with the purity of
99.5% was obtained from Hangzhou Electrochemical-Gas
Factory (Hangzhou, China). The other chemicals such as
ethanol were used as received.
2.2 SFE
SFE was carried out on the extraction system (model
HA220-50-06, Hua An SFE Company, Jiangsu, China)
which consists of a 5 L volume extractor, two 1 L separa-
tors, two syringe pumps (model 100 DX, Jiangsu, China),
and a syphonated CO2 cylinder. The operation pressure
and temperature could reach up to 50 MPa and 808C. The
extraction pressure of the system was adjusted by a pres-
sure regulator and the temperature was controlled by a
thermostatic water bath.
An orthogonal array design (OAD) test, L9 (3)4, was
employed as a chemometric method for investigating
SFE of tanshinones (dihydrotanshinone I, cryptotanshi-
none, tanshinone I, and tanshinone IIA) from the dried
root of S. miltiorrhiza. As listed in Table 1, four processing
parameters, namely, extraction pressure (P), temperature
(T), dynamic extraction time (t), and flow rate of CO2 (F),
were studied and optimized.
For each extraction, about 1.5 kg of the prepared plant
material was loaded into the extractor. After statically
extracting for 0.5 h, the sample was subjected to
dynamic extraction by supercritical CO2. The extracted
substances were collected at the bottom of separators
and were then evaporated at 608C in vacuum condition.
www.jss-journal.com
J. Sep. Sci. 2008, 31, 321 – 328
2.3 Solvent extraction (SE)
In order to compare the extraction efficiency of SFE with
conventional extraction using organic solvent, SE using
methanol was carried out according to the Chinese phar-
macopoeia (2005). Mashed Danshen (0.3 g) and 50 mL of
methanol were placed in a conical flask and the suspen-
sions were boiled for 1 h. The obtained solution was then
centrifuged at 10 000 rpm for 5 min. After filtering
through a 0.45 lm membrane filter, the supernatant sol-
ution was injected into the HPLC system for analysis.
2.4 HPLC analysis
All samples were analyzed by an Agilent HPLC (1100 Ser-
ies, Agilent, USA) equipped with a quaternary pump, a
vacuum degasser, an autosampler, a UV detector, a col-
umn heater – cooler, and a ChemStation software system.
A Lichrospher C18 column (250 mm64.6 mm id, 5.0 lm
particle size) purchased from Hanbang Science & Tech-
nology (Jiangsu, China) coupled with an Agilent C18 pre-
column (4 mm65 mm) was used at 308C. The mobile
i 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Liquid Chromatography 323
Figure 2. HPLC chromatograms
of standard mixture (A) and the
SFE products of S. miltiorrhiza
(B): 1, dihydrotanshinone I; 2,
cryptotanshinone; 3, tanshinone
I; 4, tanshinone IIA.
phase consisted of A (water) and B (ACN) flowing at
1.0 mL/min. The gradient elution of mobile phase was
60% (B) from 0 to 17 min, 60 – 80% (B) from 17 to 25 min,
80% (B) from 25 to 30 min. The injection volume was
20 lL and the detection was carried out at 270 nm for all
samples analyzed.
Representative HPLC-UV chromatograms of the four
reference compounds and the supercritical fluids
extracts of S. miltiorrhiza under the condition listed above
are shown in Fig. 2.
2.5 Preparation of standard solution
The reference standards for the investigated compounds,
i. e., dihydrotanshinone I, cryptotanshinone, tanshinone
I, and tanshinone IIA, were accurately weighed and dis-
solved in methanol to prepare a stock standard solution.
This solution was stored away from light at 48C. Quantifi-
cation was performed using external calibration. A lin-
ear response with a correlation coefficient more than
0.999 (n = 6) was obtained for the standards.
www.jss-journal.com
324 L. Wang et al. J. Sep. Sci. 2008, 31, 321 – 328

Table 1. Factors and levels of the OAD
Level P (MPa) T (8C) t (h) F (L/h)
A B C D
1 20 40 1 20
2 30 50 2 25
3 40 60 3 30
2.6 Preparation of sample solution
The product of SFE (2 mg) was accurately weighed, then
put into a clean dry volumetric flask (25 mL) and dis-
solved with methanol. The obtained solution was then
centrifuged at 10 000 rpm for 5 min. After filtering
through a 0.45 lm membrane filter, the supernatant was
injected into the HPLC system.
3 Results and discussion
3.1 SFE experimental data analysis
The first step in the SFE is to optimize the operating con-
ditions to obtain an efficient extraction of the target
compounds and avoid the coextraction of undesired
compounds such as fatty acids and their esters.
Since various parameters potentially affect the extrac-
tion process, the optimization of the experimental condi-
tions is a critical step in developing an SFE method. Based
on the previous knowledge of SFE, the extraction pres-
sure and temperature, the dynamic extraction time and
flow rate of CO2 are usually considered as the most
important factors of SFE. They were investigated at first
during this study using a three-level OAD. We focused on
the main effects of them and the interactions among
different variables were not considered in the matrix.
The extract obtained from each test was quantitatively
analyzed by HPLC to determine the contents of four tan-
shinones and then the extraction yields of these com-
pounds were calculated. The experimental results are
listed in Table 2.
Table 2 indicates that the maximum extraction yields
of the four tanshinones were 0.98, 4.90, 2.28, and
7.19 mg/g, respectively. The yields of tanshinones
obtained were the average of three repeated measure-
ments. And it was also noticed that each of them
imparted different influence upon the yields of different
tanshinones. Therefore, if the analysis is only made based
on the statistics listed in Table 2, it was hard to select the
best extraction conditions. So a further analysis was sub-
sequently performed and listed in Table 3.
From Table 3, it could be inferred that the Factor D
(flow rate of CO2) is the most significant factor according
to the R values, while the Factor C (the dynamic extrac-
tion time) is the insignificant one compared with the
others. Figure 3 was also helpful to obtain the optimized
SFE conditions. It shows the relationship between the
extraction yields of tanshinones and the four variables,
including the extraction pressure (20 – 40 MPa), extrac-
tion temperature (40 – 608C), dynamic extraction time
(1 – 3 h), and flow rate of CO2 (20 – 30 L/h).
Table 4 shows the analysis of variance (ANOVA) of the
experimental results, where the flow rate of CO2 contrib-
utes as a significant factor for yield of dihydrotanshi-
none I with 95% confidence, while temperature, pres-
sure, and flow rate of CO2 are statistically significant for
cryptotanshinone. Besides, all the four factors have no
significant effect on the yields of tanshinone I and tanshi-
none IIA. Since the four compounds have the different
chemical structures and physical properties, it is not sur-
prising that, with 95% confidence, the various param-
eters have quite different effects on the yields of the four
tanshinones. The optimization of the experimental con-
ditions should be obtained by analyzing the effects of the
different operating conditions on the yields of tanshi-
nones.

Table 2. Experimental results of the orthogonal test

Trial Parameters Extraction yielda) (mg/g) SD (n = 3)

no.
A B C D Dihydrotanshinone I Cryptotanshinone Tanshinone I Tanshinone IIA

1 1 1 1 1 0.32 l 0.01 1.28 l 0.02 0.48 l 0.04 1.74 l 0.09

2 1 2 2 2 0.98 l 0.07 4.90 l 0.28 1.24 l 0.12 4.94 l 0.42
3 1 3 3 3 0.30 l 0.02 1.24 l 0.11 0.47 l 0.03 1.77 l 0.09
4 2 1 2 3 0.15 l 0.02 0.61 l 0.04 0.24 l 0.02 0.85 l 0.05
5 2 2 3 1 0.30 l 0.02 1.15 l 0.07 0.64 l 0.05 2.00 l 0.13
6 2 3 1 2 0.63 l 0.04 2.60 l 0.13 0.76 l 0.05 2.98 l 0.07
7 3 1 3 2 0.84 l 0.05 4.39 l 0.40 2.28 l 0.17 7.19 l 0.08
8 3 2 1 3 0.70 l 0.05 2.30 l 0.21 2.10 l 0.07 5.89 l 0.09
9 3 3 2 1 0.23 l 0.02 0.75 l 0.03 0.63 l 0.04 1.41 l 0.07
a)
Extraction yield (mg/g) = the amount of the target compounds/sample mass.

i 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.jss-journal.com

J. Sep. Sci. 2008, 31, 321 – 328 Liquid Chromatography 325

3.2 Effect of extraction pressure

Pressure is a critical experimental variable for SFE as it
affects the density of the supercritical fluid, which is
directly correlated to solute's solubility. The higher
extraction pressure, at constant temperature, leads to a
higher fluid density, increasing the solubility of the com-
pounds. In this assay, the influence of pressure on the
extraction efficiency of tanshinones was studied under
different conditions by changing the pressure from 20 to
40 MPa. As shown in Fig. 3a, the increase of the extrac-
tion pressure from 20 to 30 MPa had minimal effects on
the extraction efficiencies of most tanshinones. And the
extraction efficiency of the four tanshinones was
increased markedly when pressure was increased from
30 to 40 MPa, mostly because the density of the supercrit-
ical fluid increased with higher solvating power. Taking
all of the results into consideration, within the ranges of
the parameters studied, 40 MPa was selected as the opti-
mal extraction.

3.3 Effect of extraction temperature

The effect of the extraction temperature is demonstrated
in Fig. 3b. In this work, temperatures at three different
values (40, 50, and 608C) were evaluated to optimize the
extraction process. As shown in Fig. 3b, the yields of tan-
shinones increased when increasing the extraction tem-
perature from 40 to 508C. But when the temperature
increased from 50 to 608C, the efficiency of the tanshi-
nones decreased. So the highest extraction yield was
obtained when temperature was at 508C. This maybe
resulted from two opposite effects of temperature. On
one hand, the increasing of temperature leads the sol-
vent density to decrease, reducing the solubility of sol-
utes in the supercritical fluid. On the other hand, the sol-
ubility of most organic compounds could be conspicu-
ously improved at the same time as the solute's vapor
pressure increases with increase in temperature.

3.4 Effect of dynamic extraction time

In order to obtain high yields of tanshinones, an impor-
tant extraction step of static extraction (30 min) was per-
formed. This step could make a better penetration of the
fluid into the matrix compared with the only dynamic
extraction mode. This step was followed by a dynamic
extraction to enhance solubility in the supercritical
fluid. To evaluate the effect of dynamic extraction time
on SFE of tanshinones, extraction times were adopted as
1, 2, and 3 h. Figure 3c shows that the extraction yields of
tanshinones did not increase significantly in the exten- Figure 3. Effects of (a) pressure, (b) temperature, (c)
sion of extraction time. Besides, the dynamic extraction dynamic extraction time, and (d) CO2 flow rate on SFE yields
time is found to be insignificant to influence the extrac- of tanshinones from Danshen. G, Dihydrotanshinone I; H,
tion yields rather than the other three factors according tanshinone I; F, cryptotanshinone; 6, tanshinone IIA.

i 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.jss-journal.com

326 L. Wang et al. J. Sep. Sci. 2008, 31, 321 – 328

Table 3. Analysis of L9 (3)4 test results Table 4. ANOVA table for the SFE of tanshinones by using
OA9 (34)
Parameters
Source of Sum of Degree of Mean F-value F0.05
A B C D variance squares freedom square

Dihydrotanshinone I ANOVA for the SFE of dihydrotanshinone I

M1a) 0.53 l 0.03 0.43 l 0.06 0.55 l 0.09 0.28 l 0.02 P 0.086 2 0.043 5.733 19
M2 0.36 l 0.06 0.66 l 0.05 0.45 l 0.04 0.82 l 0.11 T 0.127 2 0.064 8.467 19
M3 0.59 l 0.09 0.39 l 0.03 0.48 l 0.07 0.38 l 0.04 T 0.015 2 0.007 Pooled 19
Rjb) 0.23 0.27 0.10 0.54 F 0.482 2 0.241 32.133 19
Pooled error 0.015 2 0.007
Cryptotanshinone Total 0.71
M1 2.48 l 0.14 2.09 l 0.16 2.06 l 0.12 1.06 l 0.04
M2 1.45 l 0.10 2.78 l 0.19 2.09 l 0.12 3.96 l 0.27 ANOVA for the SFE of cryptotanshinone
M3 2.48 l 0.22 1.53 l 0.09 2.26 l 0.20 1.38 l 0.12 P 2.107 2 1.054 29.493 19
Rj 1.03 1.25 0.20 2.90 T 2.356 2 1.178 33.296 19
T 0.071 2 0.036 Pooled 19
Tanshinone I F 15.167 2 7.584 213.94 19
M1 0.73 l 0.07 1.00 l 0.08 1.12 l 0.13 0.58 l 0.12 Pooled error 0.071 2 0.036
M2 0.55 l 0.07 1.33 l 0.08 0.70 l 0.06 1.43 l 0.04 Total 19.7
M3 1.67 l 0.13 0.62 l 0.04 1.13 l 0.04 0.94 l 0.11
Rj 1.12 0.71 0.43 0.85 ANOVA for the SFE of tanshinone I
P 2.181 2 1.090 6.226 19
Tanshinone IIA T 0.751 2 0.376 2.143 19
M1 2.82 l 0.23 3.26 l 0.07 3.54 l 0.08 1.72 l 0.10 T 0.353 2 0.176 Pooled 19
M2 1.95 l 0.08 4.28 l 0.25 2.40 l 0.21 5.04 l 0.22 F 1.087 2 0.544 3.074 19
M3 4.83 l 0.08 2.05 l 0.08 3.65 l 0.10 2.84 l 0.08 Pooled error 0.353 2 0.176
Rj 2.89 2.22 1.25 3.32 Total 4.372
a)
Mi: The mean values of yields for the factors at each ANOVA for the SFE of tanshinone IIA
level. P 13.148 2 6.574 4.572 19
b)
Rj = Mi, max – Mi, min. T 7.433 2 3.716 2.584 19
T 2.884 2 1.442 Pooled 19
to the R values in Table 3. Therefore, 1 h could be selected F 17.1 2 8.55 5.952 19
Pooled error 2.884 2 1.442
as the optimal dynamic extraction time.
Total 40.565

3.5 Effect of flow rate of CO2

the results were compared. Both of them were performed
The effect of flow rate of CO2 on SFE of tanshinones is
thrice under the same condition. Figure 4 shows the
shown in Fig. 3d. It was inferred that extraction yields of
extraction yields obtained by both methods. Tested tan-
all tanshinones increased firstly with increase in CO2
shinones from left to right in this figure are in the order
flow rate from 20 to 25 L/h and attained the maximum
of their retention times in the chromatography. It can be
value near 25 L/h, then decreased with CO2 rate increas-
inferred that, in general, overall extraction yields of tan-
ing from 25 to 30 L/h. Although higher flow rate of CO2
shinones using SFE were much higher than those using
generally elevates the mass transfer between CO2 and the
SE. The RSDs of the extraction using SFE for all the target
solute by increasing the number of CO2 molecules per
compounds were less than 15%. For cryptotanshinone
unit volume to enter the extractor, there are some disad-
and tanshinone IIA, SFE shows extremely higher extrac-
vantages to extract the solute. Many researchers have
tion yields than the SE method. This may be due to some
found that too fast flow of CO2 reduces the contact time
degradation under the high temperature during SE proc-
between CO2 and the solute, especially for the solute
esses. And it also indicates that SFE is more suitable to
which has low solubility or hardly diffuses from the
extract the tanshinones from Danshen.
material matrix [25, 26]. At the same time, higher flow
rates may reduce the trapping efficiency of the system,
resulting in a lower recovery of tanshinones at higher
3.7 Validation of the chromatographic method
flow rates due to the less efficient trapping.
The method specificity has been demonstrated by suc-
cessfully separating the four investigated compounds
3.6 Comparison with SE
and the others in the SFE products. The limits of quanti-
The optimized SFE and SE techniques were employed sep- tation (LOQs) for the four investigated compounds were
arately for the extraction of tanshinones in Danshen and determined to be 14.1 ng for dihydrotanshinone I,

i 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.jss-journal.com

J. Sep. Sci. 2008, 31, 321 – 328
15.0 ng for cryptotanshinone, 20.0 ng for tanshinone I,
and 7.8 ng for tanshinone IIA at a S/N of 10, respectively.
The intra- and interday precision was determined by a
standard mixture solution of the four tanshinones under
the selected chromatography conditions six times in a
day for intraday precision and once a day on three con-
secutive days for interday precision. And the RSD was
taken as a measure of the intra- and interday precisions,
which were less than 1.26 and 0.29%, respectively.
Standard addition test was performed to determine
recoveries of the four tanshinones. In this assay, the
standard solutions with different concentration levels
were prepared and added to the samples with known
content. The resultant samples were analyzed using the
developed HPLC method mentioned above. The experi-
ments were repeated three times for each level. The ratio
of measured and added amounts was used to calculate
the recovery. The recovery of the investigated compo-
nents ranged from 97.2 to 103.8%, and their RSD values
were less than 3.7%. Considering the results of the recov-
ery test, the method is accurate.
4 Concluding remarks
Because the therapeutic effect of traditional Chinese
medicine is frequently based on the combination of the
major components, study on extraction focusing on mul-
ticomponent objectives is more reasonable to evaluate
the SFE process developed. In the present study, the
effects of four main experimental factors, extraction
pressure and temperature, dynamic extraction time, and
CO2 flow rate on the SFE yields of tanshinones from
Danshen, the dried root of S. miltiorrhiza, were investi-
gated. Statistical analysis showed that the optimum con-
ditions for extraction of tanshinones by SFE are as:
P = 40 MPa, T = 508C, and t = 1 h, and CO2 flow rate,
F = 25 L/h.
Meanwhile, a novel HPLC method was developed for
simultaneous determination of dihydrotanshinone I,
i 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Liquid Chromatography 327
Figure 4. Comparison of the
extraction yields of tanshinones
from Danshen between SFE and
SE. 1, Dihydrotanshinone I; 2,
cryptotanshinone; 3, tanshinone
I; 4, tanshinone IIA.
cryptotanshinone, tanshinone I, and tanshinone IIA from
the product of SFE with good sensitivity, precision, and
repeatability. And it can be applied as an improved qual-
ity control analysis method for Danshen products.
Financial supports from Chinese National Key Technologies R&D
Program (Grant No. 2006BAI06A08) and from the Science and
Technology Department of Zhejiang Province, China (Grant No.
2007C23G2090017) are gratefully acknowledged.
5 References
[1] Lu, Y. R., Foo, L. Y., Phytochemistry 2002, 59, 117 – 140.
[2] Kasimu, R., Tanaka, K., Tezuka, Y., Gong, Z. N., Li, J. X., Basnet, P.,
Namba, T., Kadota, S., Chem. Pharm. Bull. 1998, 46, 500 – 504.
[3] Zheng, G. C., Kakisawa, H., Pharm. Chin. J. 1989, 24, 6 – 10.
[4] Kong, D. Y., Chin. J. Pharm. 1989, 20, 279 – 285.
[5] Huang, M. H., Lu, X. Z., Chin. J. Pharm. 1992, 23, 279 – 283.
[6] Lei, X. L., Chiou, G. C., Am. J. Chin. Med. 1986, 14, 145 – 52.
[7] Tian, G. L., Zhang, T. Y., Zhang, Y. B., Ito, Y., J. Chromatogr. A 2002,
945, 281 – 285.
[8] Chang, H. M., Cheng, K. P., Choang, T. F., Chow, H. F., Chui, K. Y.,
Hon, P. M., Tan, F. W. L., Yang, Y., Zhong, Z. P., J. Org. Chem. 1990,
55, 3537 – 3543.
[9] Sun, L., Lee, H. K., J. Chromatogr. A 2003, 1014, 165 – 177.
[10] Sonsuzer, S., Sahin, S., Yilmaz, L., J. Supercrit. Fluids 2004, 30,
189 – 199.
[11] Ghasemi, E., Yamini, Y., Bahramifar, N., Sefidkon, F., J. Food Eng.
2007, 79, 306 – 311.
[12] Hawthorne, S. B., Anal. Chem. 1990, 62, 633A – 642A.
[13] Langenfeld, J. J., Hawthorne, S. B., Miller, D. J., Pawliszyn, J., Anal.
Chem. 1994, 66, 909 – 916.
[14] Taylor, L. T., Supercritical Fluid Extraction, John Wiley & Sons, Inc.,
Toronto 1996.
[15] Lang, Q., Wai, C. M., Talanta 2001, 53, 771 – 782.
[16] Dean, J. R., Liu, B., Price, R., J. Chromatogr. A 1998, 799, 343 – 348.
[17] Liu, A. H., Lin, Y. H., Yang, M., Guo, H., Guan, S. H., Sun, J. H.,
Guo, D. A., J. Chromatogr. B 2007, 846, 32 – 41.
[18] Zhao, Y., Chen, B., Yao, S. Z., J. Pharm. Biomed. Anal. 2005, 38, 564 –
570.
[19] Gu, M., Zhang, G. F., Su, Z. G., Ouyang, F., J. Chromatogr. A 2004,
1041, 239 – 243.
[20] Tian, G. L., Zhang, Y. B., Zhang, T. Y., Yang, F. Q., Ito, Y., J. Chroma-
togr. A 2000, 904, 107 – 111.
www.jss-journal.com
328 L. Wang et al. J. Sep. Sci. 2008, 31, 321 – 328

[21] Shi, Z. H., He, J. T., Yao, T. T., Chang, W. B., Zhao, M. P., J. Pharm. [25] Valle, J. M., Rivera, O., Mattea, M., Ruetsch, L., Daghero, J., Flores,
Biomed. Anal. 2005, 37, 481 – 486. A., J. Supercrit. Fluids 2004, 31, 159 – 174.
[22] Pan, X. J., Niu, G. G., Liu, H. Z., J. Chromatogr. A 2001, 922, 371 – [26] Ge, F. H., Zhang, J. C., Chen, L., Li, J., Huang, X. F., Chin. J. Pharm.
375. 2000, 31, 250 – 253.
[23] Pan, X. J., Niu, G. G., Liu, H. Z., Biochem. Eng. J. 2002, 12, 71 – 77.
[24] Zhang, J., Wang, L., Yuan, C. L., Wang, X. Q., Yu, Z. L., Chin. J. Anal.
Chem. 2005, 33, 355 – 358.

i 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.jss-journal.com