Talanta 99 (2012) 932–938

Contents lists available at SciVerse ScienceDirect

Talanta
journal homepage: www.elsevier.com/locate/talanta

Microwave-assisted extraction in combination with capillary electrophoresis

for rapid determination of isoquinoline alkaloids in Chelidonium majus L.
Qiuhong Zhou, Youping Liu, Xin Wang, Xin Di n
School of Pharmacy, Shenyang Pharmaceutical University, 103 Wenhua Road, Shenyang 110016, PR China

a r t i c l e i n f o a b s t r a c t

Article history: A simple and rapid method based on microwave-assisted extraction (MAE) followed by capillary
Received 10 April 2012 electrophoresis (CE) was developed for the quantification of eight isoquinoline alkaloids in Chelidonium
Received in revised form majus L. (Ch. majus). The key parameters affecting CE separation and MAE extraction were investigated
17 July 2012
and optimized. Complete separation of eight alkaloids was achieved within only 9 min using a 500 mM
Accepted 25 July 2012
Available online 1 August 2012
Tris–H3PO4 buffer (pH 2.5) containing 50% (v/v) methanol and 2 mM HP-b-cyclodextrin. The optimal
MAE extraction was performed at 60 1C for 5 min with methanol–water–HCl (90:10:0.5, v/v/v) as the
Keywords: extracting solvent, which gave much higher extraction efficiency in significantly shorter time than
Microwave-assisted extraction conventional heat reflux extraction (HRE) and ultrasonic extraction (USE) methods. Good linearities
Capillary electrophoresis
were obtained for all the alkaloids investigated with correlation coefficients above 0.9994. The
Chelidonium majus L.
repeatability and intermediate precision were less than 4.11% and the recoveries ranged from 98.0%
Isoquinoline alkaloids
to 103.9%. The developed method was successfully applied to 14 Ch. majus samples obtained from
different regions of China. Compared with previously reported methods, the present method offers a
dramatic savings in overall analysis time and considerable reduction in solvent consumption.
& 2012 Elsevier B.V. All rights reserved.

1. Introduction advantages including reduced run time, increased peak capacities

and less solvent consumption over conventional HPLC, it required
Chelidonium majus L. (Ch. majus), also known by the common tedious sample preparation (heat reflux extraction followed by
names ‘‘greater celandine’’ and ‘‘tetterwort’’, is one of the most solid-phase extraction) prior to UPLC analysis.
important medicinal plants of the family Papaveraceae. Modern Capillary electrophoresis (CE) is one of the most powerful tools
research has revealed that Ch. majus mainly contains a range of for natural products analysis [12–14]. With the advantages of
isoquinoline alkaloids, including sanguinarine, chelidonine, coptisine, high separation efficiency, short analysis time and low reagent
protopine, stylopine, chelerythrine, berberine and allocryptopine, etc. consumption, CE has been the focus of attention for developing
(Fig. 1) [1]. Due to the presence of these alkaloids, the extracts of new analytical methodology. Several papers have reported the
Ch. majus have been reported to possess various pharmacological use of CE to determine the main alkaloids in Ch. majus [1–3,15].
activities, such as anti-inflammatory, antimicrobial, antiviral, anti- Kulp et al. [16] has most recently developed a CE method for the
tumoral and cytotoxic effects [2–4]. Consequently, isoquinoline determination of seven isoquinoline alkaloids in Ch. majus with
alkaloids have been considered as indices for estimation of quality ultraviolet light-emitting diode-induced native fluorescence (UV-
of Ch. majus. LEDIF) detection. In spite of the enhanced selectivity and sensi-
Some high-performance liquid chromatography (HPLC) methods tivity offered by the combination of CE and UV-LEDIF, application
[5–10] have previously been reported for the determination of of this method is hampered by laborious sample preparation
isoquinoline alkaloids in Ch. majus. However, these methods suf- (ultrasonic extraction for 6 times followed by centrifugation,
fered from several disadvantages, such as time-consuming sample evaporation and reconstitution) and long analysis time.
preparation steps, large solvent consumption and long analysis time. In recent years, microwave-assisted extraction (MAE) has been
Recently, Gu et al. [11] developed an ultra-performance liquid proven to be a powerful sample extraction technique due to its
chromatography (UPLC) method for the quantification of seven ability to reduce the volume of extraction solvents and extraction
main alkaloids in Ch. majus. Although this method showed some time, improve the reproducibility and recovery of analytes and
increase sample throughput [17–20]. MAE has been successfully
applied to the extraction of some bioactive constituents in tradi-
n
Corresponding author. Tel.: þ86 24 2398 6342; fax: þ 86 24 2390 2539. tional Chinese medicines (TCMs) [21–24]. In the present work, the
E-mail address: [email protected] (X. Di). superiority of MAE over conventional extraction techniques for the

0039-9140/$ - see front matter & 2012 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.talanta.2012.07.061
 Q. Zhou et al. / Talanta 99 (2012) 932–938 933

Fig. 1. Chemical structures of eight isoquinoline alkaloids from Ch. majus.

extraction of isoquinoline alkaloids from Ch. majus was investigated 2.2. Standard solutions
and demonstrated. For the first time, a simple and rapid method
based on MAE followed by CE with photodiode-array detection was A mixed stock solution of SAN, BER, COP, CHE, CHD, PRO, ALL
developed for the simultaneous determination of eight isoquinoline and STY was prepared in methanol. A series of working standard
alkaloids in Ch. majus. The analytical advantages of the present solutions were prepared by successive dilution of the stock
method over previously published methods in terms of extraction solution with methanol. All the solutions were stored at 4 1C
time, analysis time and solvent consumption were illustrated. until use.

2.3. Extraction
2. Experimental
2.3.1. Microwave-assisted extraction
2.1. Reagents and materials MAE was performed with an Ethos A Microwave-assisted
Extraction System (Milestone, Italy). 1.0 g of dried Ch. majus
Sanguinarine (SAN), berberine (BER) and coptisine (COP) were powder (20 mesh) was weighed accurately into a 100-mL Teflon
purchased from Chengdu Mansite Pharmaceutical Co., Ltd. (Sichuan, extraction vessel and then extracted with 20 mL of methanol–
China). Chelerythrine (CHE), chelidonine (CHD), protopine (PRO) and water–HCl (90:10:0.5, v/v/v) for 5 min at 60 1C. After cooling, the
allocryptopine (ALL) were obtained from Shenzhen Medherb extract was centrifuged at 12,000 rpm for 3 min and the super-
Biotechnology Co., Ltd. (Shenzhen, China). Stylopine (STY) was natant was directly injected into the CE system.
isolated and purified from Ch. majus in the Laboratory of Drug
Metabolism and Pharmacokinetics, Shenyang Pharmaceutical Uni-
2.3.2. Ultrasonic extraction
versity (Shenyang, China) and its chemical structure was confirmed
1.0 g of dried Ch. majus powder (20 mesh) was weighed
based on UV, MS [m/z 324.1158([MþH] þ )], 1H-NMR [600 MHz,
accurately into a flask and then extracted with 20 mL of
CDCl3: d6.71(1H, s, H-1), 6.63(1H, s, H-4), 2.60–3.11(4H, m, H-5, 6),
methanol–water–HCl (90:10:0.5, v/v/v) in an ultrasonic bath for
3.51(1H, d, J¼15.3 Hz, H8-a), 4.14(1H, d, J¼15.3 Hz, H8-b), 6.65(1H,
30 min. After cooling, the extract was centrifuged at 12,000 rpm
d, J¼8.0 Hz, H-11), 6.63(1H, d, J¼8.0 Hz, H-12), 2.83(1H, dd, J¼15.6,
for 3 min and the supernatant was directly injected into the CE
11.5 Hz, H13-a), 3.21(1H, dd, J¼ 15.6, 11.5 Hz, H13-b), 3.59(1H, br.d,
system.
J¼11.5 Hz, H-14), 5.94, 5.92(each 1H, d, J¼1.4 Hz, 2,3-OCH2O-),
5.91(2H, s, 9, 10 OCH2O-)] and 13C-NMR [150 MHz, CDCl3: 104.1
(C-1), 145.2, 145.0(C-2, 3), 107.4(C-4), 126.4(C-4a), 29.3(C-5), 50.2 2.3.3. Heat reflux extraction
(C-6), 51.8(C-8), 116.2(C-8a), 141.3(C-9), 144.0(C-10), 106.5(C-11), 1.0 g of dried Ch. majus power (20 mesh) was weighed
121.3(C-12), 130.2(C-12a), 36.3(C-13), 59.9(C-14), 127.1(C-14a), accurately into a 50-mL round bottom flask and then extracted
100.9(2, 3-OCH2O-), 101.6(9, 10-OCH2O-)] data. The purities of all with 20 mL of methanol–water–HCl (90:10:0.5, v/v/v) under
the compounds were over 98% determined by HPLC-UV. Sodium reflux for 1 h in a water bath at 60 1C. After cooling, the extract
hydroxide, sodium borate, hydrochloric acid and phosphoric acid of was centrifuged at 12,000 rpm for 3 min and the supernatant was
analytical grade were obtained from Dongxing Corporation (She- directly injected into the CE system.
nyang, China), Bodi Corporation (Tianjin, China), Xinyang Corpora-
tion (Henan, China) and Damao Corporation (Tianjin, China), 2.4. CE analysis
respectively. Methanol, acetonitrile and isopropanol of HPLC grade
were supplied by Concord Corporation (Tianjin, China). Tris CE analyses were carried out using an Agilent HP3D Capillary
(499.8%) was from Shanghai Sangon Biological Engineering Tech- Electrophoresis System (Agilent Technologies, Waldbrom, Germany)
nology & Services Co., Ltd. (Shanghai, China). Samples of Ch. majus equipped with a built-in photodiode-array detector and controlled
were collected from different regions in China and were identified by the ChemStation CE software. Separation was performed on an
by Professor Qishi Sun, School of Traditional Chinese Materia uncoated fused-silica capillary (Yongnian Optic Fiber Plant, Hebei,
Medica, Shenyang Pharmaceutical University (Shenyang, China). China) of 50 mm i.d. and 375 mm o.d. with a total length of 35 cm
The voucher specimens of these samples were deposited in the (26.7 cm length to detection window). The temperature of the
Laboratory of Drug Metabolism and Pharmacokinetics, Shenyang capillary cassette was maintained at 20 1C. The running buffer was
Pharmaceutical University (Shenyang, China). 500 mM Tris–H3PO4 buffer (pH 2.5) containing 50% (v/v) methanol
934 Q. Zhou et al. / Talanta 99 (2012) 932–938

and 2 mM HP-b-cyclodextrin. The separation voltage was held at

20 kV. The detection wavelength was set at 205 nm. Samples were
loaded by pressure injection at 50 mbar for 5 s. Every day before the
first run, the capillary was consecutively flushed with 0.1 M NaOH
for 10 min, water for 10 min and the running buffer for 10 min.
Between runs, the capillary was rinsed with 0.1 M NaOH, water and
the running buffer for 3 min each.

3. Results and discussion

3.1. Optimization of CE conditions

Kulp et al. [16] reported that complete separation of seven

isoquinoline alkaloids in Ch. majus was achieved within 22 min
using an acidic phosphate buffer. In the present study, the
possibility for achieving even faster CE separation of eight iso-
quinoline alkaloids was investigated. According to the theory of
electrophoresis, it is possible to obtain increases both in speed and
efficiency by increasing the voltage [25]. At the same time, the
effects of Joule heating, which increases with increasing voltage, on
efficiency should be considered. Several types of buffers including
borate buffer, phosphate buffer and Tris–H3PO4 buffer were
attempted in the separation of isoquinoline alkaloids. Finally,
Tris–H3PO4 buffer was chosen as the running buffer, which has
lower conductivity, allowing higher voltages and better separation
efficiency. In the following study, the effects of buffer pH and
concentration, organic additive and HP-b-cyclodextrin concentra-
tion on separation efficiency were investigated.

3.1.1. Effects of buffer pH and concentration

We initially investigated the migration behaviors of eight
Fig. 2. Effect of buffer pH on the resolution of adjacent peak pairs. U1, U2 and U3
alkaloids in the range of pH 2.0–6.0. It was found that SAN, COP
are unknown components in Ch. majus.
and CHE had much higher migration velocities than others,
mainly due to their relatively strong basic properties [16]. At a
pH higher than 2.7, severe peak tailing was observed for SAN and methanol could not give satisfactory resolution, peak shape and
CHE. This phenomenon has been explained by the pH-dependen- migration time. But when a very high methanol percentage was
cies of SAN and CHE solubility [26]. Therefore, a pH optimization used, precipitation of the buffer appeared. Finally, 50% methanol
was carried out within the range of 2.0–2.7. Fig. 2 shows the effect was chosen as the optimum organic modifier.
of pH on the resolution of adjacent peak pairs. It can be seen that Using a 500 mM Tris–H3PO4 buffer (pH 2.5) containing 50%
peak pairs of CHE-U1 (U1 is a later-migrating peak near to CHE), methanol, complete separation of the eight alkaloids in a standard
BER-U2 (U2 is a later-migrating peak near to BER) and ALL-STY are mixture was achieved. However, such a condition was not
the three critical pairs that are difficult to separate. The resolution suitable for real application, since the peak pair of CHE-U1 was
of peak pair of ALL-STY increased with the increase of pH, but not well separated. To improve the resolution of this peak
meanwhile the resolution of peak pairs of CHE-U1 and BER-U2 pair, addition of different concentrations of HP-b-cyclodextrin
decreased. The compromising result was achieved at pH 2.5. (0.5–5 mM) to the buffer was attempted. The results showed that
Subsequently, the concentration of Tris–H3PO4 buffer in the range the resolution of peak pair of CHE-U1 increased with the increase
of 200–500 mM was optimized at the optimum buffer pH of 2.5. of HP-b-cyclodextrin concentration, but meanwhile the resolu-
With the increase of buffer concentration, the migration times tion of peak pair of COP-CHE decreased. The best compromise was
were prolonged and the separation efficiency was increased. obtained with 2 mM HP-b-cyclodextrin.
When the buffer concentration was higher than 500 mM, exces- Under the above optimum conditions, electropherograms of a
sive current was observed. As a result, 500 mM of Tris–H3PO4 standard mixture of eight alkaloids (A) and an extract from a
buffer was chosen as the optimum buffer concentration. Ch. majus sample (B) are shown in Fig. 3.

3.1.2. Effects of organic additive and HP-b-cyclodextrin

3.2. Optimization of MAE conditions
concentration
As most of the isoquinoline alkaloids in Ch. majus have low
The effects of extraction solvent, temperature, time and
solubility in aqueous media, addition of an organic modifier to the
solvent to material ratio on the extraction efficiency of eight
running buffer has been reported to increase the solubility and
alkaloids were investigated systematically under different MAE
enhance the separation efficiency. Different organic additives includ-
conditions, followed by CE analysis [27]. The areas of CE peaks
ing methanol, acetonitrile and isopropanol were studied. Precipita-
were used for the comparison of MAE results.
tion was found when a high percentage of acetonitrile was added to
500 mM Tris–H3PO4 buffer. Addition of isopropanol had a negative
effect on the separation efficiency. So methanol was chosen as an 3.2.1. Selection of solvent in MAE
organic additive and its percentage was optimized in the range Selection of an appropriate solvent is fundamental to obtaining
of 40% to 60%. It was found that a relatively low percentage of an optimal MAE condition [20]. For the present work, different
 Q. Zhou et al. / Talanta 99 (2012) 932–938 935

Fig. 3. Electropherograms of a standard mixture of eight alkaloids (A) and an Fig. 4. Effect of temperature on the extraction of eight alkaloids from Ch. majus
extract from a Ch. majus sample (B). Peaks: (1) sanguinarine; (2) coptisine; (n¼3).
(3) chelerythrine; (4) berberine; (5) chelidonine; (6) protopine; (7) allocryptopine;
(8) stylopine; U1, U2 and U3 are unknown components in Ch. majus. Buffer:
500 mM Tris–H3PO4 buffer (pH 2.5) containing 50% (v/v) methanol and 2 mM HP-
b-cyclodextrin. Separation capillary: 50 mm i.d. uncoated fused-silica capillary,
35 cm in length (26.7 cm effective length). Applied voltage: 20 kV. Detection
wavelength: 205 nm.

extraction solvents including methanol, ethanol, water, methanol–

water (90:10, 70:30, 50:50, v/v) and ethanol–water (90:10, 70:30,
50:50, v/v) were tested with MAE under the same condition. It was
found that methanol–water (90:10, v/v) gave the highest extraction
efficiency. As the isoquinoline alkaloids are weak bases, the addition
of small quantities of acid to the extraction solvent can improve the
solubility of alkaloids [28]. Different percentages of hydrochloric
acid (0, 0.5%, 1%, 2%) in methanol–water (90:10, v/v) were further
tested. The results showed that adding some acid to the solvent
could significantly enhance the extract efficiency of eight alkaloids.
But the percentage of hydrochloric acid in the solvent was found to
have negligible effect on the MAE results. Hence, methanol–water–
HCl (90:10:0.5, v/v/v) was selected for further experiments. Fig. 5. Effect of solvent to material ratio on the extraction of eight alkaloids from
Ch. majus (n ¼3).

3.2.2. Effects of extraction temperature and time

Extraction temperature is a key factor in the optimization of MAE
reached the highest at 20 mL/g. Then most of the peak areas
procedure [29]. In order to investigate the effect of temperature on
decreased slightly, perhaps due to inadequate stirring of the solvent
the extraction efficiency of eight alkaloids from Ch. majus, four
by microwaves. Therefore, 20 mL/g was considered as the optimal
different temperatures ranging from 40 to 100 1C were examined. As
ratio of solvent to material for the MAE process.
shown in Fig. 4, the extraction efficiency of the alkaloids was
Based on the above findings, the optimal MAE condition for the
improved at elevated temperatures since increasing temperature
extraction of eight isoquinoline alkaloids in Ch. majus was: 1.0 g
will enhance solvent diffusivity, reduce surface tension and increase
of sample, 20 mL of methanol–water–HCl (90:10:0.5, v/v/v), 60 1C
solute solubility. However, at temperatures over 60 1C, a slight
and 5 min. It should be noted that repeated extraction did not
decrease of some peak areas were observed, perhaps as a result of
give significantly higher extraction efficiency than a single extrac-
thermal degradation of some alkaloids. Hence, an extraction tem-
tion. So in view of saving time and energy, one-step extraction
perature of 60 1C was selected for further experiments.
was enough to extract the alkaloids from Ch. majus.
The effect of extraction time on extraction efficiency was
The optimal MAE method was compared with two conven-
studied while maintaining the extraction temperature at 60 1C.
tional extraction methods, namely HRE and USE. The results are
Extraction times of 5, 10 and 30 min were examined. The results
presented in Fig. 6. It can be seen that MAE gave significantly
showed that no significant increase in peak area was obtained
higher extraction efficiency than HRE and USE. Moreover, MAE
with the increase of the extraction time (figure not shown).
took much less extraction time (only 5 min) than HRE (1 h) and
Therefore, 5 min was chosen as the optimal time for MAE.
USE (30 min). With the capability of parallel processing of multi-
ple sample (10 samples in a single batch), MAE is much more
3.2.3. Effect of solvent to material ratio attractive for high-throughput sample preparation.
Using an extraction temperature of 60 1C and an extraction time
of 5 min, the effect of solvent to material ratio on extraction 3.3. Method validation
efficiency was investigated. Extraction was carried out at four
different ratios of solvent to material (10, 15, 20 and 30 mL/g). It 3.3.1. Linearity
can be seen from Fig. 5 that the normalized peak areas of the The linearity of the method was evaluated by injecting a series
alkaloids increased with the increase of solvent to material ratio and of standard solutions in triplicate at six concentration levels.
936 Q. Zhou et al. / Talanta 99 (2012) 932–938

Calibration curves were constructed by non-weighted least 3.3.4. Accuracy

squares linear regression analysis of peak area (y) of the analyte The accuracy of the method was determined by performing
versus the nominal concentration (x). An analysis of variance spike and recovery experiments. Six replicate samples were
(ANOVA) was performed to test the significance of the regression. prepared by spiking known amounts of standards to a real
The lack-of-fit (LOF) test was used to determine whether the Ch. majus sample. Percentage recoveries were calculated by
selected model is adequate to describe the experimental data. The comparison of found and added amounts of the analytes in spiked
results of regression analysis are summarized in Table 1. Good samples. The original amounts of the analytes in the sample were
correlations were obtained for all the analytes with r values subtracted from the measured amounts of each spiked sample
higher than 0.9994. The p-values for LOF test are greater than before calculating recoveries. The recoveries of the eight alkaloids
0.05 for all the analytes, indicating that the linear regression ranged from 98.0% to 103.9% (Table 2), demonstrating good
models are adequate for the experimental data. reliability of the method for analysis of eight alkaloids in
Ch. majus.

3.3.2. Limit of detection (LOD) and limit of quantification (LOQ)

The LOD and LOQ were estimated as 3 and 10 times the signal- 3.3.5. Stability
to-noise ratio (S/N), respectively. The LOQs of eight analytes were The stability of the analytes in the final extract stored at room
2.15–14.15 mg/mL and the LODs were 0.65–4.28 mg/mL (Table 1). temperature was investigated by replicate injections of a freshly
Although the LODs for some alkaloids are an order of magnitude prepared sample solution at 2-h interval up to 8 h. The RSDs of
higher than those determined by CE with UV-LEDIF detection, the the assay results were 2.44–4.42%, which indicate that the sample
limits are low enough to determine the presence of all the solution is stable at room temperature for at least 8 h.
alkaloids under consideration in real Ch. majus samples.
3.3.6. Robustness
The robustness of the developed method was evaluated by
3.3.3. Precision
making deliberate variations in CE parameters such as capillary
The injection precision was determined by 10 subsequent
temperature, buffer pH, Tris concentration and separation voltage.
injections of a standard solution. The relative standard deviations
The parameters were varied 5% below and above the value set in
(RSDs) of migration time and peak area were in the range of
the method. Resolutions of peak pairs of COP-CHE, CHE-U1, BER-
0.23–0.54% and 0.95–3.11%, respectively. The repeatability of the
U2 and ALL-STY (four critical pairs) were used as indicators. No
method was assessed by six replicate analyses of the same Ch.
significant change was observed through variations of capillary
majus sample. The RSDs of migration time and peak area were all
temperature and separation voltage. Variations in buffer pH and
below 4.11% (Table 2). Intermediate precision was examined
Tris concentration resulted in variations in resolution of about
similarly to that of repeatability but by two analysts, on 3 con-
15%, indicating that these two parameters need to be closely
secutive days. As shown in Table 2, the RSDs ranged from 0.20% to
controlled. The effect of capillary on resolution was studied using
4.00%. The results indicate an acceptable level of precision.
two different batches of capillaries. No significant variations were
observed regarding the resolution of four critical pairs.

3.4. Real sample analysis

The proposed analytical method was applied to analysis of 14

batches of Ch. majus samples obtained from different regions of
China. The quantitative results were summarized in Table 3. It can
be seen that COP was the most abundant alkaloids in all Ch. majus
samples. The contents of ALL and CHE were significantly lower than
those of the other alkaloids in all samples. Due to the differences in
the region of cultivation and harvesting time, the contents of eight
alkaloids in Ch. majus varied greatly from sample to sample.

3.5. Comparison with previously reported methods

Comparison of the proposed method with some published

analytical methods for the determination of isoquinoline alkaloids
in Ch. majus was summarized in Table 4. It can be seen that the
Fig. 6. Comparison of MAE with conventional extraction techniques (n¼3). MAE approach significantly reduced the sample preparation time

Table 1
Linearity, LOD and LOQ for quantification of eight alkaloids in Ch. majus.

Compound Linear range (mg/mL) Regression equation r FLOF LOQ (mg/mL) LOD (mg/mL)

Sanguinarine 14.1–450.0 y¼ 0.397x þ 1.224 0.9996 0.343(p4 0.05) 14.10 4.27

Coptisine 28.3–904.0 y¼ 0.702x þ8.491 0.9995 1.325(p 40.05) 14.15 4.28
Chelerythrine 6.2–198.4 y¼ 0.617x þ 9.840 0.9998 1.969(p 40.05) 6.20 1.87
Berberine 12.5–400.0 y¼ 2.063x  8.823 0.9994 2.863(p 40.05) 12.50 3.78
Chelidonine 22.8–728.0 y¼ 3.658x  0.586 1.0000 1.740(p4 0.05) 2.28 0.69
Protopine 12.3–394.0 y¼ 4.168x  2.013 0.9996 3.761(p 40.05) 2.50 0.75
Allocryptopine 5.0–160.8 y¼ 9.083x  11.14 0.9994 4.011(p4 0.05) 2.15 0.65
Stylopine 12.3–394.0 y¼ 6.185x  6.210 0.9995 0.928(p4 0.05) 2.40 0.90
 Q. Zhou et al. / Talanta 99 (2012) 932–938 937

Table 2
Precision and accuracy for quantification of eight alkaloids in Ch. majus.

Compound Injection precision RSD (%) Repeatability RSD (%) Intermediate precision RSD (%) Recovery (%)

Peak area Migration time Peak area Migration time Peak area Migration time

Sanguinarine 2.05 0.23 2.91 0.20 4.00 0.21 99.9 73.0

Coptisine 2.20 0.25 2.23 0.21 2.86 0.23 100.2 72.9
Chelerythrine 2.49 0.25 4.11 0.21 3.06 0.20 99.5 74.8
Berberine 2.47 0.26 2.88 0.21 3.17 0.22 101.8 74.1
Chelidonine 3.11 0.24 2.24 0.22 3.30 0.23 100.7 73.4
Protopine 1.50 0.25 1.74 0.21 2.53 0.23 98.0 72.8
Allocryptopine 0.95 0.26 3.57 0.21 2.92 0.22 103.9 73.7
Stylopine 1.17 0.54 3.46 0.30 2.65 0.28 100.0 72.7

Table 3
Contents of eight alkaloids (mg/kg) in 14 batches of Ch. majus samples (n¼ 3).

Sample Sanguinarine Coptisine Chelerythrine Berberine Chelidonine Protopine Allocryptopine Stylopine

Liaoning1 10807 80 5768 7 305 338 7 13 1161 7 50 864 7 4 833 7 38 1577 7 23477 102
Hebei2 1122 7 19 2933 7 312 384 7 2 517 7 5 977 7 12 511 7 3 1107 4 11277 17
Heilongjiang1 11017 10 3524 7 25 437 7 13 505 7 9 798 7 19 522 7 3 1507 9 20267 104
Sichuan1 17607 23 8799 7 375 783 7 21 902 7 25 10087 27 876 7 10 2087 6 11507 37
Liaoning2 9097 2 4138 7 9 294 7 1 908 7 2 687 7 4 678 7 4 1207 5 17967 23
Zhejiang 1127 7 39 67007 218 476 7 20 500 7 21 719 7 14 599 7 6 1757 4 11597 61
Jilin1 1391 7 31 74087 1103 459 7 4 682 7 14 885 7 19 7007 23 2097 10 1118 7 18
Sichuan2 1456 7 42 80457 50 635 7 15 634 7 24 851 7 35 748 7 20 1947 10 1617 7 68
Heilongjiang2 1183 7 14 28047 61 445 7 12 531 7 10 886 7 11 448 7 4 1277 2 20697 80
Shanxi 1224 7 40 6227 7 122 420 7 2 541 7 4 722 7 37 566 7 20 1507 3 18737 3
Jiangsu 8527 6 4974 7 186 298 7 3 410 7 7 514 7 12 475 7 9 1347 3 12757 18
Hernan 1144 7 16 5988 7 229 403 7 9 501 7 14 6707 30 560 7 24 1657 3 17047 24
Jilin2 1319 7 12 60297 145 426 7 16 550 7 2 793 7 29 634 7 24 1577 2 16277 13
Hebei1 12097 41 32097 120 432 7 16 499 7 4 1012 7 36 542 7 11 1277 6 16747 27

Table 4
Comparison of different methods for quantification of alkaloids in Ch. majus.

Sample preparation Sample analysis References

Solvent Solvent
Extraction Subsequent sample Number of
Time consumption Method Time consumption
methoda treatment analytes
(mL/g sample) (mL per run)

MAE Centrifugation 8 min 20 CE 9 min 8 o 10  5 Current

IE and USE Laying aside 12.5 h 10 HPLC 50 min 8 50 [6]
USE Centrifugation, evaporation, SPE 1.5 h 600 HPLC 40 min 5 40 [9]
IE Evaporation 4 360 HPLC 25 min 7 12.5 [10]
weeks
HRE SPE, filtration 1.5 h 30 UPLC 20 min 7 8 [11]
USE Centrifugation, evaporation, 1h 240 CE 24 min 7 o 10  5 [16]
reconstitution

a
MAE, microwave-assisted extraction; IE, immerse extraction; USE, ultrasonic extraction; SPE, solid-phase extraction; HRE, heat reflux extraction.

from several hours to 8 min and the analysis time of the present CE in sample analysis time and solvent consumption. The combina-
method was 2–6 times shorter than those reported in the literature. tion of MAE with CE was an effective method for rapid analysis of
Moreover, considerable reduction of solvent consumption was isoquinoline alkaloids in Ch. majus.
achieved using the present method. These results indicate that
MAE combined with CE analysis is a good choice for rapid extraction
and analysis of isoquinoline alkaloids in Ch. majus. Acknowledgments

The present study was supported by a Key Project for Drug

4. Conclusions Innovation grant from the Ministry of Science and Technology of
China (2009ZX09301-012).
A simple, fast and reliable method based on MAE combined
with CE analysis was developed and validated for the simulta-
neous determination of eight isoquinoline alkaloids in Ch. majus References
for the first time. The developed method was successfully applied
to determine eight alkaloids in 14 batches of Ch. majus obtained [1] S. Sturm, H. Stuppner, Electrophoresis 19 (1998) 3026–3032.
[2] Q. Liu, Y.J. Liu, M.L. Guo, X.B. Luo, S.Z. Yao, Talanta 70 (2006) 202–207.
from different regions of China. In comparison with previously [3] P.G. Pietta, P.L. Mauri, C. Gardana, M.L. Colombo, F. Tome , Phytochem. Anal. 6
published methods, the present method offers a dramatic savings (1995) 196–202.
938 Q. Zhou et al. / Talanta 99 (2012) 932–938

[4] M. Then, K. Szentmihályi, Á. Sárkozi, V. Illés, E. Forgács, J. Chromatogr. A 889 [17] W.H. Xiao, L.J. Han, B. Shi, Sep. Purif. Technol. 62 (2008) 614–618.
(2000) 69–74. [18] F.L. Hu, C.H. Deng, Y. Liu, X.M. Zhang, Talanta 77 (2009) 1299–1303.
[5] C. Bugatti, M.L. Colombo, F. Tome , J. Chromatogr. 393 (1987) 312–316. [19] Z.B. Li, D.N. Huang, Z.X. Tang, C.H. Deng, X.M. Zhang, Talanta 82 (2010)
[6] C.Q. Niu, L.Y. He, J. Chromatogr. 542 (1991) 193–199. 1181–1185.
[7] L.F. Han, W. Nowicky, V. Gutmann, J. Chromatogr. 543 (1991) 123–128. [20] S.N. Tan, J.W.H. Yong, C.C. Teo, L. Ge, Y.W. Chan, Ch.S. Hew, Talanta 83 (2011)
[8] Á Sárközi, G. Janicsák, L. Kursinszki, Á Kéry, Chromatographia 63 (2006) 891–898.
S81–S86. [21] G.F. Barbero, M. Palma, C.G. Barroso, Anal. Chim. Acta 578 (2006) 227–233.
[9] L. Kursinszki, Á. Sárközi, Á. Kéry, É. Szöke, Chromatographia 63 (2006) [22] M. Careri, C. Corradini, L. Elviri, A. Mangia, J. Chromatogr. A 1152 (2007)
S131–S135. 274–279.
[10] J. Suchomelová, H. Bochořáková, H. Paulová, P. Musil, E. Táborská, J. Pharm. [23] W.Y. Ma, Y.B. Lu, R.L. Hu, J.H. Chen, Z.Z. Zhang, Y.J. Pan, Talanta 80 (2010)
Biomed. Anal. 44 (2007) 283–287. 1292–1297.
[11] Y. Gu, D.W. Qian, J.A. Duan, Z.Z. Wang, J.M. Guo, Y.P. Tang, S. Guo, J. Sep. Sci. [24] Y.P. Li, G.K. Skouroumounis, G.M. Elsey, D.K. Taylor, Food Chem. 129 (2011)
33 (2010) 1004–1009. 570–576.
[12] K. Tian, Y.S. Wang, Y.L. Chen, X.G. Chen, Z.D. Hu, Talanta 72 (2007) 587–593. [25] R.T. Kennedy, I. German, J.E. Thompson, S.R. Witowski, Chem. Rev. 99 (1999)
[13] S. Zhang, S.Q. Dong, L.Z. Chi, P.G. He, Q.J. Wang, Y.Z. Fang, Talanta 76 (2008) 3081–3131.
780–784. [26] R. Vespalec, M. Vlčková, V. Kubán, Electrophoresis 26 (2005) 3265–3272.
[14] J.Y. Pan, S. Zhang, L.S. Yan, J.D. Tai, Q. Xiao, K. Zou, Y. Zhou, J. Wu, [27] V. Camel, Trends Anal. Chem. 19 (2000) 229–248.
J. Chromatogr. A 1185 (2008) 117–129. [28] Y. Shen, C. Han, Y.X. Jiang, X.J. Zhou, Z.O. Zhu, X.X. Lei, Talanta 84 (2011)
[15] H. Stuppner, M. Ganzera, J. Chromatogr. A 717 (1995) 271–277.
1026–1031.
[16] M. Kulp, O. Bragina, P. Kogerman, M. Kaljurand, J. Chromatogr. A 1218 (2011)
[29] R.M. Smith, J. Chromatogr. A 975 (2002) 31–46.
5298–5304.