Content uploaded by Changqing Yi
Author content
All content in this area was uploaded by Changqing Yi
Content may be subject to copyright.
Microchim. Acta 147, 237–243 (2004)
DOI 10.1007/s00604-004-0238-y
Original Paper
High-Performance Liquid Chromatographic Determination
of Quinolizidine Alkaloids in Radix Sophora Flavescens Using
Tris(2,20-Bipyridyl)Ruthenium(II) Electrochemiluminescence
Changqing Yi1, Peiwei Li2, Ying Tao1,and Xi Chen1;
1The Key Laboratory of Analytical Sciences of the Ministry of Education and the Department of Chemistry,
Xiamen University, Xiamen 361005, China
2Department of Biology, Indiana University, Bloomingtion, IN 47401, USA
Received November 10, 2003; accepted March 1, 2004; published online June 21, 2004
#Springer-Verlag 2004
Abstract. Electrogenerated chemiluminescences
(ECLs) of quinolizidine alkaloids including matrine
(MT), sophocarpine (SC), and sophoridine (SRI) are
studied. The light emission is caused by an electro-
oxidation reaction between RuðbpyÞ3
2þand the ter-
tiary amino group on the alkaloid compounds. A
thin-layer flow cell equipped with a glassy carbon disk
electrode (22.1 mm
2
) at the potential of þ1.30 V (vs.
Ag=AgCl) was applied for ECL observation. MT, SC
and SRI were separated and quantitatively determined
within 25 min by an ODS-80 Ts reversed-phase
column with a mobile phase containing 80 mmol L1
NaH
2
PO
4
–K
2
HPO
4
buffer þacetonitrile (7:3) þ
40 mmol L1sodium dodecyl sulfate (pH 6.5). The
determination limit at an S=N of 3 ranged from 3
109gmL
1for MT, 6 109gmL
1for SC and
1109gmL
1for SRI. The recoveries are from 92
to 108%, with repeatability ranging from 1.3 to 4.5%
(relative standard deviation). The method was success-
fully applied to the determination of quinolizidine
alkaloids in Sophora flavescens samples.
Key words: Electrochemiluminescence; HPLC; quinolizidine
alkaloids; tris(2,20-bipyridyl)ruthenium.
Sophora viciifolia is a bush that grows throughout
the southwest of China and is an important source of
Chinese medicine. The root of sophora viciifolia is a
common traditional Chinese herbal drug used to treat
fever, cystitis, haematuria and edemas [1]. The herb is
known to contain quinolizidine alkaloids as bioactive
constituents [2, 3]. A previous study has shown that
sophoridine is the main alkaloid in this herb [4].
Recently, several chromatographic separation meth-
ods, such as high-performance liquid chromatography
(HPLC) [5], capillary electrophoresis (CE) [6] and
thin-layer chromatography (TLC) [7], have been ap-
plied to analyze sophora viciifolia for the presence
of quinolizidine alkaloids. However, there are certain
limitations for these analytical methods. For example,
TLC lacks quantitative precision, while UV-detection
of HPLC has an unsatisfactory determination limit,
and the CE methods are not conveniently used in an
aqueous system.
Chemiluminescence analysis using RuðbpyÞ3
2þ
has become an attractive detection means for bio-
chemical substances including amines, amino acids
or bioactive alkaloids [8–10] because of its low
detection limit and wide linear working range, with rel-
atively simple instrumentation. Generally, the electro-
generated chemiluminescence (ECL) of RuðbpyÞ3
2þ
Author for correspondence. E-mail: xichen@xmu.edu.cn
is due to the reaction of its oxidized state,
RuðbpyÞ3
3þ, with a reductant to give RuðbpyÞ3
2þas
follows [11]:
RuðbpyÞ3
2þ!RuðbpyÞ3
3þe
RuðbpyÞ3
3þþR02NCH2R!RuðbpyÞ3
2þþR02NþCH2R
RuðbpyÞ3
2þþR02NþCH2RþH2O!R02NH þOCHR
þRuðbpyÞ3
þþ2Hþ
RuðbpyÞ3
þþRuðbpyÞ3
3þ!RuðbpyÞ3
2þþRuðbpyÞ3
2þ
RuðbpyÞ3
2þ!RuðbpyÞ3
2þþhðmax ¼610 nmÞ
In the course of the ECL study, we found that
matrine (MT), sophocarpine (SC), and sophoridine
(SRI) produced strong luminescence in the presence
of RuðbpyÞ3
3þ. In this paper, we describe a new
method for HPLC-ECL determination of these alka-
loids. The factors influencing the ECL intensity of
the alkaloids are discussed. The procedure is then
applied to the determination of MT, SC and SRI in
Sophora flavescens samples without a derivatization
procedure.
Experimental
Chemicals and Standard Solution
Ru(bpy)
3
Cl
2
6H
2
O was obtained from Sigma Chemical Company
(St. Louis, Mo, USA) and was used without further purification.
Matrine, sophocarpine and sophoridine were purchased from the
National Institute for Control of Pharmaceutical and Biological Prod-
ucts (Beijing, China). All other chemicals were of analytical reagent
grade. Pure water from a Simplicity Personal Ultrapure Water Sys-
tem (Millipore, USA) was used to prepare all buffers and other
solutions which were all filtered through 0.45 mm membrane filter
film and degassed in an ultrasonic bath before use. The stock solu-
tions (0.1 mg mL1) of MT, SC and SRI were prepared with metha-
nol, and the working solutions were diluted by carrier solutions.
Instrumentation and Procedure
The instrumentation for HPLC experiments used in this paper has
been described previously [12]. A three-electrode system was used
for potentiostatic control of the electrolytic system. The working
electrode was a glassy carbon disk (GC, 22.1 mm
2
), and the reference
electrode was Ag=AgCl. The three-electrode system was composed
in a thin-layer electrolytic cell that was designed in our laboratory for
ECL observation. The main body of the cell was set by two pieces of
Teflon block tightly fixed to each other. Since luminescence intensity
is closely related to the orifice shape and thickness of the Teflon
Fig. 1. Structural formula of three quinolizi-
dine alkaloids
Fig. 2. Experimental setup for ECL-HPLC
238 C. Yi et al.
spacer inside the cell, a spacer sheet with a thickness of 50 mmwas
chosen. The volume of the thin-layer cell was 1.5mL.
HPLC was performed using an Agilent 1100 (Aglient Co. Ltd.,
USA) liquid chromatographer equipped with a Rheodyne 7125
sample injector (Cototi, CA, USA, 20 mL) and an ODS-80 Ts
reversed-phase column (150 4.6 mm, Tosoh, Japan). The mobile
phase was 80 mmolL1NaH
2
PO
4
–K
2
HPO
4
buffer solution (pH
6.0) and CH
3
CN (7:3) containing 40mmol L1sodium dodecyl
sulfate (SDS). The flow rate was kept at 0.3 mL min1. The reagent
solution was prepared by dissolving 0.8 mmol L1RuðbpyÞ3
2þ
in 0.05 mol L1NaOH–NaAc–0.3 mol L1KNO
3
buffer solution
(pH 10.0). The flow rate was kept at 0.5 mL min1.
Sample Preparation and Data Processing
0.5 g of pulverized dried sample was immersed in 10mL methanol
at room temperature for 12 hours. The samples were ultrasonically
treated for 20min, and the same procedure was repeated twice.
The extracts were combined and centrifuged at 3500rev min1
for 10 min, filtered through a 0.45 mm cellulose acetate membrane
filter, and then diluted to 50.0mL with methanol as a stored sam-
ple solution. The stored sample solution was diluted another 25
folds with the mobile phase (80mmol L1NaH
2
PO
4
–K
2
HPO
4
þ
CH
3
CN (7:3) þ40 mmol L1SDS, pH 6.0) and then used for HPLC
analysis.
Calibration curves in the concentration range of 0.01 to
20.0 mgmL
1for MT, SC and SRI were constructed by plotting
the peak area of analyte standards of MT, SC and SRI concentration
in Sophora flavescens Ait samples. Least squares linear regression
analysis was used to determine the slope, intercept and correlation
coefficient. The concentrations of MT, SC and SRI in samples were
determined from the HPLC peak areas by using the equations of
linear regression obtained from the calibration curves.
For the recovery studies, a given amount of MT, SC and SRI
standard solution was added to the crude drugs that contained
known amounts of the alkaloids for recovery testing, and the mix-
ture was extracted and analyzed using the same procedure as
described previously.
Results and Discussion
Voltammetric Analysis and Effect
of Applied Potential
The chemiluminescent reaction mechanisms between
RuðbpyÞ3
3þand oxalate [13] or amino acid [14]
have been studied. When RuðbpyÞ3
2þwas oxidized
to RuðbpyÞ3
3þat a glassy carbon electrode, the
strong reducing intermediate (radical ions) resulting
from oxidized oxalate or amino acid produces the
excited state, RuðbpyÞ3
2þ, by an electron transfer
reaction with trivalent ruthenium species. Owing
to the applied potential affecting ECL behaviors
of RuðbpyÞ3
2þand coexisting compounds, cyclic
voltammetry (CV) was performed in 0.1 mol L1
KH
2
PO
4
–NaOH buffer (pH 8.0). A reversible anodic
wave with þ1.05 V could be obtained in the pres-
ence of RuðbpyÞ3
2þon the potential scan in
the positive direction resulting from oxidation of
RuðbpyÞ3
2þto RuðbpyÞ3
3þ. From the anodic current
and the experimental results of the applied potential
shown in Fig. 3, the anodic current intensity of
RuðbpyÞ3
2þwas increased in the presence of MT, SC
or SRI, and the anodic current intensity at þ1.05 V
was changed from 14.4 mA for RuðbpyÞ3
2þto 33.7 mA
for MT, 27.4 mA for SC and 34.4 mA for SRI, respec-
tively. This evidences that oxidized RuðbpyÞ3
2þ
reacted with the quinolizidine alkaloids in alkaline
solution and that the addition of quinolizidine alka-
loids in RuðbpyÞ3
2þsolution obviously increased the
oxidation rate of RuðbpyÞ3
2þon the glassy carbon,
resulting in the increase of ECL intensity to 68 mV,
86 mV and 59 mV for SRI, MT and SC at a concen-
tration of 200 ng mL1, respectively. A further evi-
dence of the electrolytic effect on the increase of
RuðbpyÞ3
2þcould be obtained by chronocoulometry
as described by Zhao [15]. As shown in Fig. 4, eight
approximately linear sections were observed both in
the forward (F) and reverse (R) directions.
In the presence of RuðbpyÞ3
2þsolution, Q
F(Ru)
,
Q
R(Ru)
and t
1=2
represent the relationships as ex-
pressed by formula 1:
QFðRuÞ¼1:38481067:5812 105t1=2;
QRðRuÞ¼5:86001065:7864 105t1=2
ð1Þ
Fig. 3. Voltammetric response of CV experiment for three quino-
lizidine alkaloids. (a) 1.0 mM RuðbpyÞ3
2þþ0.1 M NaH
2
PO
4
–
K
2
HPO
4
buffer (pH 8.0), (b)70mgmL
1SC þ1.0 mM
RuðbpyÞ3
2þþ0.1 M NaH
2
PO
4
–K
2
HPO
4
buffer (pH 8.0), (c)
70 mgmL
1MT þ1.0 mM RuðbpyÞ3
2þþ0.1 M NaH
2
PO
4
–
K
2
HPO
4
buffer (pH 8.0), (d)70mgmL
1SRI þ1.0 mM
RuðbpyÞ3
2þþ0.1 M NaH
2
PO
4
–K
2
HPO
4
buffer (pH 8.0)
High-Performance Liquid Chromatographic Determination of Quinolizidine Alkaloids 239
In the presence of RuðbpyÞ3
2þand MT solution:
QFðMTÞ¼7:00001061:4464 104t1=2;
QRðMTÞ¼7:27931063:2274 105t1=2
ð2Þ
In the presence of RuðbpyÞ3
2þand SC solution:
QFðSCÞ¼5:42381061:1931 104t1=2;
QRðSCÞ¼6:73281063:7941 105t1=2
ð3Þ
In the presence of RuðbpyÞ3
2þand SRI solution:
QFðSRIÞ¼7:3751061:7301 104t1=2;
QRðSRIÞ¼7:92261062:1674 105t1=2
ð4Þ
The slope of the linear section indicates the effec-
tive concentration of RuðbpyÞ3
3þon the electrode in
each case. Before the reaction of luminous emission,
it could be found that in the presence of MT, SC
or SRI, the effective concentration of RuðbpyÞ3
3þis
1.96-fold, 1.57-fold and 2.23-fold compared to the
values observed in the presence of RuðbpyÞ3
2þsolu-
tion only. After luminous emission, the concentration
of RuðbpyÞ3
3þdecreased by 77.7% for MT, 68.2% for
SC and 87.5% for SRI. It is obvious that the ECL
intensity of alkaloids derives from their different elec-
trolytic efficiency on the glassy carbon electrode.
Moreover, in RuðbpyÞ3
2þ–NaOH-NaAc solution
alone, the weaker luminescence emission was caused
by electro-oxidation of RuðbpyÞ3
2þto RuðbpyÞ3
3þ,
and then by the reaction of RuðbpyÞ3
2þwith dis-
solved oxygen and other impurities in the aqueous
alkaline solution [16]. Its intensity was enhanced by
the addition of MT. The phenomenon could also be
observed in the same buffer solution of RuðbpyÞ3
2þ–
SC and RuðbpyÞ3
2þ-SRI. The maximum emission
wavelength for their RuðbpyÞ3
2þsolutions was at
610 nm, which confirmed that the luminescence in
the mixture of RuðbpyÞ3
2þwith the quinolizidine
alkaloids is due to the presence of an excited state
of RuðbpyÞ3
2þ.
The ECL intensity was measured as a function of
the applied potential at a glassy carbon electrode
(22.1 mm
2
) in a flow injection system. The volume
of sample solution (20 mL) containing 50 ng mL1
quinolizidine alkaloids was injected into a carrier
stream of 0.3 mmol L1RuðbpyÞ3
2þ, 0.05 mol L1
KH
2
PO
4
–NaOH (pH 8.0) and 0.3 mol L1KNO
3
,
then passed through the working electrode which
was held at a set of potentials linearly increasing
from þ0.80 V (vs Ag=AgCl). The ECL intensity was
clearly observed when the applied potential was over
þ1.10 V and increased as the applied potential went
up to þ1.20 V, and a potential of þ1.30 V was
selected for HPLC application.
Optimization of pH and Carrier Solution
The chemiluminescence between RuðbpyÞ3
2þand
some compounds could be obtained in a wide pH
range. Tryptophan, for instance, was at pH 3.0 [17],
oxalate at pH 6.0 [13], ascorbic acid at pH 7.2 [12]
and some amino acids at pH10 [14]. A very weak ECL
signal was gained before the pH of the carrier solution
reached 5.5. The ECL response increased when the
pH of the carrier solution was raised over 5.5, and it
maintained a constant value above pH 8.5. The suit-
able pH for the carrier solution was thus fixed at 9.0.
The luminescence intensity of RuðbpyÞ3
2þwas in-
creased by changing the supporting electrolyte in the
order NO3>CH3COOF>CH3OCl>Br
in the carrier solution. The experimental results showed
that the luminescence intensity of 0.05 mgmL
1
SRI was obtained at 86 mV, 68 mV and 62 mV in
0.2 mol L1KNO
3
,CH
3
COONa, and KCl solution, re-
spectively. The ECL intensity was also affected by dif-
ferent kinds of buffer solution at pH 9.0. Comparing
Fig. 4. Chronocoulograms of RuðbpyÞ3
2þand quinolizidine alka-
loids; initial potential: 600 mV; final potential: 1200mV; pulse
width: 0.25 s; step: 2. (a) 1.0 mM RuðbpyÞ3
2þþ0.1 M NaH
2
PO
4
–
K
2
HPO
4
buffer (pH 8.0), (b)70mgmL
1SC þ1.0 mM
RuðbpyÞ3
2þþ0.1 M NaH
2
PO
4
–K
2
HPO
4
buffer (pH 8.0), (c)
70 mgmL
1MT þ1.0 mM RuðbpyÞ3
2þþ0.1 M NaH
2
PO
4
–
K
2
HPO
4
buffer (pH 8.0), (d)70mgmL
1SRI þ1.0 mM
RuðbpyÞ3
2þþ0.1 M NaH
2
PO
4
–K
2
HPO
4
buffer (pH 8.0)
240 C. Yi et al.
the ECL intensities of SRI in NaOH–K
2
HPO
4
,
NaAc–NaOH, KHCO
3
–NaOH, Na
2
B
4
O
7
–NaOH, with
the same concentrations of 0.03 mol L1containing
0.2 mol L1KNO
3
and pH 9.0, brighter luminescence
was observed in NaOH–NaAc buffer solution.
The concentration of RuðbpyÞ3
2þaffected the
luminescence intensities of quinolizidine alkaloids.
Although a higher concentration of RuðbpyÞ3
2þ
resulted in a larger response, it also yielded much
noise, and hence 0.3 mmol L1RuðbpyÞ3
2þwas
selected. The ECL response also depended on the flow
rate of the carrier solution. The luminescence inten-
sity decreased with an increase in the flow rate over
0.50 mL min1, since the reaction of oxidized quino-
lizidine alkaloids with RuðbpyÞ3
3þoccurred as a
result of electrode reactions. The luminescence in-
tensity increased at lower flow rates, but the blank
response also increased. Consequently, an appropriate
flow rate was 0.40 mL min1.
In addition, the effect of eluent for HPLC applica-
tion was investigated. With increasing concentration
of potassium dihydrogen phosphate and sodium
hydroxide in the eluent, the luminescent intensity
increased slightly up to over 0.08 mol L1. It was
shown that the ECL intensities decreased by 22% to
33% for these three alkaloids if the concentration of
SDS in the eluent rose up to 40 mmol L1. Although
a higher concentration of SDS was helpful for the
separation of MT, SC and SRI, the ECL responses
decreased greatly at the same time. In our experiment,
methanol was used to extract the alkaloids from the
crude drugs, and the addition of methanol in the elu-
ent was considered to improve the separation of
quinolizidine alkaloids. Unfortunately, the base line
became unstable and the noise clearly increased when
the concentration of methanol was above 10%, since
methanol also contributed to ECL with RuðbpyÞ3
3þ
[18], and the baseline separation for the three quino-
lizidine alkaloids could not be obtained at this metha-
nol concentration. A series of acetonitrile solutions
Table 1. ECL intensities of substances examined by using an
80 mM NaH
2
PO
4
–K
2
HPO
4
buffer (pH 6.0) – acetonitrile (7:3)
and 40 mM sodium dodecyl sulfate
Compound Detection limit
a
(pmol)
Retention time
(min)
D-gluconic acid 15 9.43
Malic acid 5.6 11.20
Citric acid 2.8 14.85
Tartaric acid 8.2 12.45
Ascorbic acid 3.2 10.69
Saccharic acid 16 11.43
D-glucose 20 9.35
D-fructose 22 9.18
Mannose 8.5 9.49
Sucrose 14 8.84
Methanol 35 10.32
Ethanol 40 11.08
1-Proparol 56 12.36
Histidine 2.0 15.41
Hydroxyproline 0.5 10.30
Proline 0.5 9.87
a
Signal-to-noise ratio of 3, per 20 mL injection.
Fig. 5. Chromatogram for separation of the four quinolizidine
alkaloids. (1) Sophocarpine, (2) sophoridine, (3) matrine (A) mixture
of quinolizidine alkaloids; (B,C,D) samples from Guizhou
province; (E) sample from Sichuan province. Eluent: 80 mM
NaH
2
PO
4
–K
2
HPO
4
buffer (pH 6.0) þacetonitrile (7:3) þ40 mM
sodium dodecyl sulfate, flow rate: 0.3 mL min1; carrier solution:
0.8 mM RuðbpyÞ3
2þþ0.15 M NaOH– NaAc þ0.3 M KNO
3
buffer
solution (pH 10.0), flow rate: 0.5 mL min1; applied potential,
þ1.30 V (vs Ag=AgCl), glassy carbon electrode 22.1 mm
2
; separa-
tion column, ODS-80 Ts reversed-phase column (150 4.6 mm,);
sample volume, 20 mL; temperature, 50 C; concentration of each
quinolizidine alkaloid, 5 mgmL
1
Table 2. Calibration curves and detection limits (S=N¼3)
Alkaloids Regression
equations
r Detection
limits
(ng mL1)
MT y ¼0.832x þ0.732 0.9915 3.0
SC y ¼0.9193x 0.084 0.9987 6.0
SRI y ¼1.041x þ0.748 0.9944 1.0
High-Performance Liquid Chromatographic Determination of Quinolizidine Alkaloids 241
with concentrations ranging from 5% to 50% (v=v)
were added to the eluent, and their effects on separa-
tion were compared. Addition of acetonitrile made the
peaks sharper and shortened the separation times, and
30% acetonitrile was best able to improve the separa-
tion. Higher concentrations of acetonitrile would
reduce the retention time very much, but also make
the ECL baseline unstable. Therefore, 30% acetoni-
trile was selected for the separation of the alkaloids. A
higher pH of eluent increased the ECL intensity, so
pH 6.5 was chosen as the limitation of pH for the
separation column. Furthermore, the luminescence
intensity decreased with an increase of the eluent flow
rate when it was over 0.5 mL min1, because of
the reaction of oxidized quinolizidine alkaloids with
RuðbpyÞ3
3þand the dilution of RuðbpyÞ3
2þin solu-
tion. Lower eluent flow rates caused higher ECL re-
sponses and noise; meanwhile, a longer analysis time
was required. Consequently, 80 mmol L1NaH
2
PO
4
–
K
2
HPO
4
buffer (pH 6.5) þacetonitrile (7:3) þ
40 mmol L1sodium dodecyl sulfate, and an appropri-
ate flow rate of 0.30 mL min1for the eluent were used
in the following experiments.
HPLC Analysis
Interferences commonly found in lupin plants were
investigated. It is known that compounds with an
–N- group, such as amines and amino acids, will emit
light when they react with RuðbpyÞ3
3þin an alkaline
solution. Among twenty amino acids, only the second-
ary amino acids, such as histidine, proline and hy-
droxyproline, gave stronger light emission. Weaker
luminescence of carbohydrates and acids with –OH
groups as saccharic acid, malic acid, or tartaric acid
was also noted. Stronger light emissions resulted only
from a few compounds such as sodium oxalate,
proline, hydroxyproline, histidine, ascorbic acid and
citric acid under experimental conditions. Fortunately,
as shown in Table 1, the retention times of all of these
coexisting compounds are less than 15 min under the
experimental conditions.
The calibration curves of MT, SC and SRI were
constructed in the range of 0.01– 20 mgmL
1.
The regression equations of these curves and their
correlation coefficients were calculated as shown in
Table 2, together with their detection limits. Compar-
ing the retention times of standards with those of
samples and by adding standards of a given quantity
to the samples allows the peak identification of the
three alkaloids to be realized. The recoveries of the
HPLC determination for the quinolizidine alkaloids
were assessed by comparing their peak areas. Typical
HPLC-ECL chromatograms of a mixture of pure MT,
SC and SRI at a concentration of 5 mgmL
1and the
extracts of the Sophora flavescens Ait samples (under
the conditions described above) are shown in Fig. 4.
The results of quinolizidine alkaloids analysis in the
samples from different areas of China are listed in
Table 3. The recovery of the detection was 92 to
108, and the RSD is satisfactory.
In conclusion, the chemiluminescent oxidation
reactions of the three quinolizidine alkaloids (MT,
SC and SRI) with electrogenerated RuðbpyÞ3
2þcan
be applied to sensitive and reproducible detection
for the alkaloids in sophora vicilfolia samples after
HPLC separation.
Acknowledgement. We express our sincere appreciation to the
Natural Scientific Foundation of China (20375033) and the Research
Foundation of Key Laboratory of Photochemistry, Center for
Molecular Science, Institute of Chemistry, and Chinese Academy of
Sciences for their support of this study.
References
[1] Xiao P G (ed) (1993) ‘‘A pictorial encyclopedia of Chinese
medical herbs’’, Jpn edn 6, Chuokoron-Sha., Tokyo, p 95
[2] Zhang S Y, Li S Y (1993) Chin Pharm J 28: 328–330
[3] Zhao B G (1980) Acta Pharm Sinica 15: 182–183
Table 3. Concentration of quinolizidine alkaloids in radix sophorae flavescentis (n ¼5)
Sample MT (mg g1) SC (mg g1) SRI (mg g1) Spike test Recovery%
Added=mg g1Found=mg g1MT SC SRI
MT SC SRI
B 2.43 0.07 5.10 0.15 4.84 0.12 2.00 4.34 6.98 6.29 98 3.2 93 2.9 92 2.5
C 0.90 0.04 3.72 0.07 3.47 0.10 2.00 3.13 5.95 5.25 108 4.5 104 2.1 96 2.8
D 1.05 0.03 2.18 0.06 2.11 0.03 2.00 3.11 3.85 4.27 102 3.5 92 2.8 104 1.3
E 1.20 0.03 2.04 0.07 2.26 0.04 2.00 3.01 3.88 4.05 94 2.5 96 3.5 95 1.8
Sample B, C and D were from Guizhou province, China; sample E was from Sichuan province, China.
242 C. Yi et al.
[4] Wang X K, Li J S, Wei L X (1995) Zhongguo Zhongyao Zazhi
20: 168–169
[5] Liu C S, Sheng L S (1993) Chin Trad Herbal Drugs 24:
405–406
[6] Song J Z, Xu H X, Tian S J (1999) J Chromatogr A 857:
303–311
[7] Jin L X, Cui T T, Zhang G D (1993) Acta Pharm Sinica 28:
136–139
[8] Noffsinger J B, Danielson N D (1987) Anal Chem 59: 865–868
[9] Jackson W A, Bobbitt D R (1994) Anal Chim Acta 285:
309–320
[10] Li M J, Chen X, Yi C Q, Wang X R (2001) Anal Sci 17 [Suppl]
a105
[11] He L, Cox K A, Danielson N D (1990) Anal Lett 232: 195–201
[12] Chen X, Sato M (1995) Anal Sci 11: 749–754
[13] Rubinstein I, Martin C R, Bard A J (1983) Anal Chem 55:
1580–1582
[14] Brune S N, Bobbit P R (1992) Anal Chem 64: 166–170
[15] Zhao C Z, Xu S Y, Su Y, Zhao G L (2002) Analyst 127: 889–891
[16] Rubinstein I, Bard A J (1981) J Am Chem Soc 103: 512–516
[17] Vchikura K, Kirisawa M (1991) Anal Sci 7: 971–975
[18] Chen X, Sato M, Lin Y J (1998) Microchem J 58: 13– 20
High-Performance Liquid Chromatographic Determination of Quinolizidine Alkaloids 243
