Você está na página 1de 8

Food Chemistry 81 (2003) 631–638

www.elsevier.com/locate/foodchem

Analytical, Nutritional and Clinical Methods Section

Analysis of green tea catechins: comparative study between


HPLC and HPCE
Matteo Bonoli*, Marco Pelillo, Tullia Gallina Toschi, Giovanni Lercker
Dipartimento di Scienze degli Alimenti, Università di Bologna, Via Ravennate 1020, Cesena (FC), 47023, Italy

Received 11 July 2002; received in revised form 13 November 2002; accepted 13 November 2002

Abstract
A comparison between a borate–phosphate–SDS based MEKC and an RP-HPLC method for the separation of seven tea cate-
chins and gallic acid in a green tea extract is here proposed. Under optimised conditions, HPCE offered several advantages respect
to time of analysis (compounds were separated within 4.5 min), sensitivity (HPCE LODs were about 20–100 times lower than
HPLC ones) and solvent consumption. HPCE displayed excellent migration time repeatability (RSD% on MT < 2%, and RSD%
on RMT < 1%), whereas HPLC showed slightly more quantification ruggedness (total amount catechins RSD% was < 2% for
HPLC and <6% for HPCE). This study highlights the effective possibilities of application of HPCE in the food chemistry field for
routine analysis.
# 2003 Elsevier Science Ltd. All rights reserved.
Keywords: MEKC; HPLC-DAD; Catechins; Green tea; Antioxidants

1. Introduction varieties, climate and cultivation. Furthermore, for


green tea extracts, quantity of GTCs depends mainly on
Green tea, a non-fermented product obtained just by the technologies applied during extraction, concen-
leaves desiccation, is an excellent source of polyphenolic tration and preservation processes.
antioxidants, which are mainly represented by a group The quality control of these extracts must compre-
of compounds having a flavan-3-olic structure and gen- hend the determination of the total amount of catechins
erally defined as green tea catechins, GTCs. The main and the quantification of every single isomer by a rapid
types of catechins present in green tea are seven: ()- and reproducible method. In literature, as far as the
gallocatechin (GC), ()-epigallocatechin (EGC), (+)- analysis of GTCs is concerned, there are many applica-
catechin (C), ()-epigallocatechin-3-gallate (EGCG), tions by high performance liquid chromatography
()-epigallocatechin (EC), ()-gallocatechingallate (HPLC; Dalluge & Nelson, 2000; Gallina Toschi, Bor-
(GCG), ()-epicatechingallate (ECG). These substances doni, Hrelia, Bendini, Lercker, & Biagi, 2000; Horie &
show a strong antioxidant activity, and are known also Kohata, 2000; Pelillo, Biguzzi, Bendini, Gallina Toschi,
for their effects on human body, such as the anti- Vanzini, & Lercker, 2002; Sano, Tabata, Suzuki,
bacterial, antiviral and antiallergenic properties (Dal- Degawa, Miyase, & Maeda-Yamamoto, 2001; Wang,
luge & Nelson, 2000; Hertog, Freskens, Hollman, Helliwell, & You, 2000; Zuo, Chen, & Deng, 2002), but
Katan, & Kromhout, 1993; Hollman, Hertog, & Katan, there are only few examples of analysis by high-perfor-
1996; Miura et al., 2001; Sesso, Gaziano, Buring, & mance capillary electrophoresis (HPCE) (Aucamp,
Hennekens, 1999). Hara, & Apostolides, 2000; Barroso & Van de Werken,
The amount of green tea catechins varies according to 1999; Horie & Kohata, 2000; Larger, Jones, &
the type of raw material, in particular to the kind of Dacombe, 1998; Lee & Ong, 2000; Stach & Schmitz,
2001; Worth, Wielcfsler, & Schmitz, 2000). This pro-
* Corresponding author. Tel.: +39-0547-636121; fax: +39-0547-
mising analytical technique is still at a developing stage,
382348. especially if compared it to other hyphenated and
E-mail address: mbonoli@foodsci.unibo.it (M. Bonoli). already consolidated techniques, such as gas capillary
0308-8146/03/$ - see front matter # 2003 Elsevier Science Ltd. All rights reserved.
doi:10.1016/S0308-8146(02)00565-4
632 M. Bonoli et al. / Food Chemistry 81 (2003) 631–638

chromatography and liquid chromatography (Robards determination), 60% of polyphenols (expressed as


& Antolovich, 1997). HPCE is characterised by a nota- EGCG; w/w, HPLC determination), and not less than
ble separation efficiency (normally ranging from 105 to 40% of EGCG. A 1000 mg/ml GTE solution was pre-
106 theoretical plates) and sensitivity, short analysis pared by dissolving 100 mg of GTE in 100 ml of an
times, small sample volumes, low running costs and the HPLC-grade water/formic acid solution (99.7/0.3, v/v).
possibility of working mainly with water mobile phases. Samples to be analysed by MEKC were added with
This allows the lowering of laboratory costs for the sol- acetone (GTE/acetone, 99/1, v/v) as a marker, since it
vents purchase and dismantlement, and it has a positive does not interact with the micelles.
impact on the environment. For these reasons, in the last
years, the number of publications regarding the applica- 2.3. Standard preparation and method evaluation
tion of CE in food analysis has been considerably
increased (Corradini & Cavazza, 1998; Issaq, 1999; Lin- The stock solutions were prepared dissolving 0.5–1
deberg, 1996a, 1966). In addition, the flexibility of this mg of each standard compound in 1 ml of HPLC-grade
technique, which was primarily used for charged mole- water/formic acid solution (99.7/0.3, v/v). Each stock
cules of big dimensions, is now increased by various solution was then used for the preparation of the diluted
operative modes, having a wide range of applications. solutions, for the calibration curves (from 0.005 to 1000
This work proposes a comparison between a MEKC mg/ml for ECG, EC, EGC, GCG, EGCG and GA; from
and an HPLC separation and quantitation of seven 0.005 to 500 for C; from 0.005 and 200 mg/ml for GC).
catechins and gallic acid, in a purified green tea extract. For both analytical methods, each point of the cali-
The MEKC and the HPLC methods cited were previously bration curves corresponded to the mean value obtained
optimised in our laboratory (Bonoli, Gallina Toschi, & from three injections. Detection limits (LODs) were
Lercker, 2002; Bonoli, Colabufalo, Pelillo, Gallina estimated from the calibration curves (signal/noise
Toschi, & Lercker, submitted for publication; Pelillo et al., ratio, S/N, equal to 3).
2002), and this study highlights the effective possibilities of The intra-assay precision was calculated from ten
application of CE in the food chemistry field. consecutive determinations of the same GTE sample;
the inter-assay precision was calculated from five differ-
ent samples. Both statistical parameters were re-assessed
2. Materials and methods 1 week later, using the same samples.

2.1. Reagents and chemicals 2.4. Micellar electrokinetic capillary chromatography:


instrument and conditions
()-Gallocatechin (GC, 98%), (+)-catechin (C,
98%), ()-epigallocatechin (EGC, 98%), ()-epigallo- Electrophoretic analyses were performed using a
catechin-3-gallate (EGCG, 95%), ()-epicatechin (EC, Beckman P/ACE 5000 model equipped with an UV–vis
purity not specified), ()-epicatechingallate (ECG, detector (Beckman Instruments, Inc., Fullerton, CA). A
98%), ()-gallocatechingallate (GCG, purity not speci- PC, equipped with Beckman P/ACE Station software,
fied) and Gallic acid (GA, 99%) were purchased from accomplished data acquisition and processing. The
Sigma Chemical Co. (St. Louis, MO, USA). Sodium capillary cartridge contained undeactivated fused silica
dodecyl sulfate for HPCE, water for HPCE, 0.1 M tubing (50 mm i.d.375 mm o.d.), supplied from Beck-
sodium hydroxide, 1 M sodium hydroxide and 0.1 M man. Total capillary length was 47 cm, whereas effective
hydrochloric acid were from Fluka (Neu-Ulm, Switzer- length was 40 cm.
land). Potassium dihydrogen phosphate, HPLC-grade The capillary cartridge was conditioned by flushing
methanol and formic acid were purchased from Carlo with 1 M sodium hydroxide for 3 min, 0.1 M sodium
Erba Reagents (Milan, Italy). HPLC-grade water, hydroxide for 3 min, HPCE-grade water for 3 min and
HPLC-grade acetone and 50 mM sodium tetraborate finally with the running buffer for 5 min; when not in
(pH 9.3) for HPCE were from Merck (Darmstadt, Ger- use, it was stored in water to prevent buffer crystal-
many). HPLC-grade acetonitrile was supplied by Pro- lization. The running buffer was daily prepared by mix-
labo (Paris, France). All the other chemicals and solvents ing three parts of 20 mM potassium dihydrogen
were high-analytical grade. Double distilled water was phosphate (solid salt diluted in HPLC-grade water), one
prepared in our laboratory from deionized water. part of 50 mM sodium tetraborate and two parts of 200
mM sodium dodecyl sulfate (solid salt diluted in HPLC-
2.2. Sample preparation grade water). It was optimized to pH=7 by adding 0.1
M hydrochloric acid; it was then filtered through a cel-
Indena (Milan, Italy) kindly donated green tea extract lulose acetate 0.2 m syringe filter (Orange Scientific,
(GTE, GreenselectTM). This aqueous spray-dried Waterloo, Belgium) and sonicated in a ultrasonic bath
extract contains less than 0.1% of caffeine (w/w, HPLC for 10 min.
M. Bonoli et al. / Food Chemistry 81 (2003) 631–638 633

Samples were injected hydrodynamically in the anodic chins cited in this work, gallic acid and caffeine. To
end at low pressure mode (0.5 psi) for 1 s. Electro- optimize the capillary electrophoresis method, theo-
phoretic run was carried out at 30 kV for 5 min, main- bromine and theophylline were added to the standard
taining the capillary temperature at 29  C, resulting in a mixture too. At the optimized conditions, the HPCE
current of approximately 85 mA. Between each run the method permitted whole separation of those 11 com-
capillary was rinsed at high pressure (20 psi) with pounds, while two peaks, corresponding to GA and
NaOH 0.1 M for 2 min, HPCE-grade water for 2 min GC, have not been resolved by the HPLC method. In
and then re-equilibrated with the running buffer for 2 fact, the separation of GA from GC is not easy to
min. At the end of each run, the capillary cartridge was obtain, and most of works in literature does not report
also rinsed with HPCE-grade water for 2 min. All their separation (Dalluge & Nelson, 2000; Horie &
washing steps were performed at 29  C. The running Kohata, 2000), but using long elution gradients (Wang
buffer was changed every three runs. et al., 2000). Applying both optimized methods caffeine
UV detection was performed at 200 nm; rise time was was completely separated to the others compounds,
set at 0.17 s and data rate was 10 Hz. Peak identification even if it was not detected in here-analyzed extract. It is
was carried out by spiking the GTE sample with stan- important to separate caffeine to others components,
dard compounds. since it is one of the main substances in tea infusions.
Both optimised methods were employed to perform
2.5. High-performance liquid chromatography: analyses of real tea infusions, but only the here-reported
instrument and conditions GTE extract was utilized to compare these two analy-
tical methods (Bonoli et al., 2003 in press; Pelillo et al.,
HPLC analyses were performed on a HP Series 1100 2002). To date, HPLC determination of tea catechins
(Hewlett Packard, Wilmington, DE, USA), equipped have been prevalently carried out by using RP octadecyl
with a binary pump delivery system, a degaser (model silyl columns (25 cm4.6 mm i.d) (Horie & Kohata,
G1322A), an autosampler (Automatic Liquid Sampler, 2000). From a preliminary work, we have demonstrated
ALS, model G1312A), a HP diode-array UV–vis detec- that using a 3.0 mm i.d. column, the resulting improve-
tor (DAD, model G1315A); integration and data ela- ment in LODs (limits of detection) was higher than one
boration were performed by ChemStation software order of magnitude, and the retention times of the tes-
(Hewlett Packard). A column RP LunaTM 5 mm C18, 25 ted compounds, in the two types of columns, were
cm3.0 mm i.d. (Phenomenex, Torrance, CA, USA) similar (paper to be published).
with Rheodyne precolumn filter 7335 model, was used. As shown in Fig. 1, the seven catechins were sepa-
All solvents were filtered with 0.45mm Millipore filter rated within 18 min; the total analysis time was about 40
disk. The following gradient elution was carried out: min, including the 6 min for column conditioning after
mobile phase A, double distilled water/methanol/formic each run. The elution order was: GC (coeluting with
acid (74.7/25/0.3; v/v/v); mobile phase B, acetonitrile/ GA), EGC, C, EGCG, EC, GCG and ECG. Fig. 1A
formic acid (99.7/0.3; v/v). The linear gradient elution shows the separation of a standard mixture (all stan-
system was: 100% A for the first 8 min, and then dard solution were 100 mg/ml), while Fig. 1B shows the
reached the 100% B in 24 min, standing at 100% B for 6 typical chromatogram of a GTE sample (concentration:
min. The gradient returned to 100% A in 4 min, stand- 1000 mg/ml; injection volume: 1 ml). Detection limits
ing at 100% A for others 6 min as post-time. The flow ranged from 0.385 to 0.025 mg/ml, which correspond to
rate was 0.5 ml/min. The quantification of the catechins EGC and GA, respectively.
by DAD was performed at 270 nm. Of each sample 1 ml In literature, HPCE catechin separation methods are
was injected, after filtration through a 0.45-mm filter very heterogeneous, in terms of application modes
disk. (CZE, MEKC), kind of buffer (background electrolytes
Identification of compounds was carried out by com- composition, micellar agents and their concentration)
paring retention times and UV spectra of the unknown and run conditions (applied voltage, capillary tempera-
peaks to those of the standards. ture). This fact may be due to the relatively recent
introduction of this technique in food analysis.
Fig. 2 shows the HPCE-electropherograms of the
3. Results and discussion same samples (standard mixture and GTE solution,
respectively). The main advantage of here developed
3.1. HPLC and MEKC analyses capillary electrophoresis method, respect to the HPLC
method, was the whole separation of the eight com-
As previously cited, the optimization of both methods pounds each others, in a very short time (within 4.5
was previously performed in our laboratory (Bonoli et min); as earlier cited, the separation of GA from other
al., 2002, submitted for publication; Pelillo et al., 2002), polyphenolic species is not easy to perform, especially
with a standard mixture constituted by the seven cate- from GC. Furthermore, the HPLC GA–LOD is about
634
M. Bonoli et al. / Food Chemistry 81 (2003) 631–638
Fig. 1. (A) HPLC-chromatogram of pure standards (all standard solutions were 100 mg/ml); (B) HPLC-chromatogram of GTE sample (concentration: 1000 mg/ml; injection volume: 1 ml). UV
detection was performed at 270 nm; the other analytical conditions are described in the methodology section. Peak identification: GC, ()-gallocatechin; EGC, ()-epigallocatechin; C, (+)-catechin;
EGCG, ()-epigallocatechin-3-gallate; EC, ()-epigallocatechin; GCG, ()-gallocatechingallate; ECG, ()-epicatechingallate; GA, gallic acid.
M. Bonoli et al. / Food Chemistry 81 (2003) 631–638
Fig. 2. (A) HPCE-electropherogram of the standard mixture (C, 500 mg/ml; GC, 200 mg/ml; EGC, 1000 mg/ml; EGCG, 1000 mg/ml; GCG, 1000 mg/ml; ECG, 1000 mg/ml; EC 1000; GA, 1000 mg/ml);
(B) HPCE-electropherogram of GTE sample. UV detection was performed at 200 nm; the other analytical conditions are described in the methodology section. Peak identification: GC, ()-gallo-
catechin; EGC, ()-epigallocatechin; C, (+)-catechin; EGCG, ()-epigallocatechin-3-gallate; EC, ()-epigallocatechin; GCG, ()-gallocatechingallate; ECG, ()-epicatechingallate; GA, gallic acid;
Ac, acetone (EOF marker).

635
636 M. Bonoli et al. / Food Chemistry 81 (2003) 631–638

20 times higher than the HPCE GA–LOD. Although correlation coefficients (r) of the calibration curves (cal-
the RSD% for GA quantification, by HPCE, was culated using peak areas) are not significantly different,
somehow high (> 30%), the fact that it was possible to while HPCE-LODs are 20–100 times lower than those
detect this compound by this technique is quite impor- obtained by HPLC. Probably, we could justify lower
tant from the qualitative standpoint. LODs of the MECK method respect to the HPLC
The slight difference between migration times of method mentioning the higher sensitivity in the detec-
the components of the standard mixture and those tion of catechins at 200 nm and the higher signal-to-
of the GTE samples (see Fig. 2A and B) could be noise ratio at this wavelength (Lee & Ong, 2000).
due to the different analytes concentration Unfortunately, it was not possible to use 200 nm as
employed. detection wavelength in HPLC due to the interference
The elution order is listed as follows: GC, C, EGC, of the mobile phase.
EGCG, GCG, ECG, EC and GA. These eight com- Relative standard deviations of HPLC retention times
pounds were completely separated within 4.5 min; a whole ranged from 0.20 to 4.14%, whereas average RSD% of
analysis required about 12 min (rinse steps included) and HPCE migration times varied between 0.17 and 1.78%.
no sample clean-up pre-treatment was necessary. Considering relative migration times (RMT) with ace-
tone as marker, the RSDs% did not reach 1% (max.
3.2. HPLC and MEKC performances value was 0.89%). In terms of quantification, the results
obtained by HPLC method displayed lower variability
Table 1 shows performance parameters (calibration than those of HPCE: RSDs% of total amounts of cate-
parameters, intra- and inter-assay precision), of GTE chins were 0.77–1.72% and 1.01–5.54% for HPLC and
samples of both HPLC and MEKC methods. The HPCE, respectively (Table 1). Nevertheless, data varia-

Table 1
HPLC and HPCE calibration curves parameters and repeatability study on quantification (values are expressed in mg of catechins for 100 mg of
extract)

Analytes r Areasa LOD (mg/ml)b First quantification (n=5) After 1 week quantification (n=5)

INTRA-assay INTER-assay INTRA-assay INTER-assay

MeanS.D. %RSD MeanS.D. %RSD MeanS.D. %RSD Mean S.D. %RSD

GC-HPLC 0.995 0.166 4.23 0.15 3.58 4.400.13 2.95 4.550.11 2.50 4.580.10 2.10
GC-HPCE 0.997 0.0013 1.21 0.05 4.38 1.160.11 9.62 1.160.08 7.26 1.190.04 3.07

C-HPLC 0.999 0.224 0.83 0.02 2.59 0.800.06 7.06 0.890.08 8.70 0.860.03 3.53
C-HPCE 0.992 0.0012 0.75 0.02 2.38 0.720.04 5.17 0.770.03 3.32 0.760.01 1.39

EGC-HPLC 0.999 0.385 11.82 0.15 1.27 11.920.16 1.38 12.070.12 0.96 12.130.26 2.17
EGC-HPCE 0.999 0.0017 10.34 0.21 2.06 10.740.77 7.13 10.080.89 8.79 10.680.35 3.23

EGCG-HPLC 0.999 0.192 38.78 0.40 1.02 39.070.30 0.77 40.300.50 1.24 40.290.67 1.67
EGCG-HPCE 0.999 0.0018 34.46 0.51 1.47 34.521.95 5.66 36.651.91 5.20 37.161.12 3.03

GCG-HPLC 0.997 0.084 0.69 0.03 4.35 0.720.03 3.94 0.850.04 4.56 0.790.05 6.60
GCG-HPCE 0.994 0.0033 1.10 0.11 10.12 1.190.11 9.36 1.260.15 11.69 1.240.08 6.44

ECG-HPLC 0.998 0.040 10.10 0.16 1.56 10.040.09 0.87 10.550.30 2.80 10.310.15 1.47
ECG-HPCE 0.999 0.0023 8.77 0.24 2.77 8.810.30 3.37 9.520.49 5.13 9.740.51 5.28

EC-HPLC 0.999 0.321 5.55 0.14 2.57 5.570.12 2.22 6.210.33 5.37 5.910.24 4.11
EC-HPCE 0.999 0.0023 5.07 0.14 2.70 4.920.29 5.82 5.330.12 2.27 5.470.13 2.29

GA-HPLC 0.999 0.0250 NDc – NDc – NDc – NDc –


GA-HPCE 0.999 0.0051 0.06 0.02 31.49 0.090.03 31.07 0.100.04 37.14 0.110.04 35.92

TOT-HPLC 71.98 0.56 0.77 72.520.64 0.88 75.400.73 0.96 74.881.29 1.72
TOT-HPCE 61.77 0.62 1.01 62.153.45 5.54 64.382.72 4.22 66.341.91 2.88
a
Correlation coefficients of the calibration curves using peak area.
b
For S/N=3
c
Not detectable.
M. Bonoli et al. / Food Chemistry 81 (2003) 631–638 637

Table 2 4. Conclusions
Differences in total amount of catechins and gallic acid (TACs) found
by HPLC and HPCE
Two different analytical techniques (MEKC and
 a
d S:D:b t0 Significativityc HPLC) for the analysis of green tea catechins, were
compared. This study highlights the effective possibility
Total amount of catechins 10.04 1.06 18.94 S
of application of HPCE in the food chemistry field. The
Pn
a ðTACsHPLC;i TACsMEKC;i Þ MEKC method here suggested show higher sensitivity,
d ¼ i¼1 ; n ¼ 4:
b
n
Standard deviation (n =4).
resolution, efficiency and migration times repeatability
c
S=significant; NS=not significant. than the HPLC method, even if the latter displayed a
slightly better repeatability in the quantification of total
amount of catechins. It should be pointed out that
HPCE LODs were about 20–100 times lower than
HPLC ones, which is extremely advantageous when
bility in HPCE is highly acceptable for a hyphenated analyzing real systems, such as food products. More-
analytical technique. As you can see in Table 1, no GA over, HPCE is very convenient in terms of analysis time
data have been reported in HPLC quantification. Only (12 min for total analysis against 40 min for the HPLC
GC quantification data have been tabulated because, one) and solvents consumption for routine analysis of
from further mass spectrometry analyses, GA was not tea extracts.
detected (Pelillo et al., 2002). This evidence confirmed
that the sensitivity of capillary electrophoresis permitted
a better qualitative investigation of this extract than References
liquid chromatography.
Regarding elution time, HPCE is much faster than
Aucamp, J. P., Hara, Y., & Apostolides, Z. (2000). Simultaneous
HPLC, since total analysis time for the eight compo- analysis of tea catechins, caffeine, gallic acid, theanine and ascorbic
nents is 12 min for this MEKC method whereas HPLC acid by micellar electrokinetic capillary chromatography. Journal of
separation is achieved in 40 min. Table 2 shows the sta- Chromatography A, 876, 235–242.
tistical interpretation of results. The following formula Barroso, M. B., & Van de Werken, G. (1999). HRC-determination of
green and black tea composition by capillary-electrophoresis. Jour-
was applied:
nal of High Resolution Chromatography, 22, 225–230.
  Bonoli, M., Colabufalo, P., Pelillo, M., Gallina Toschi, T., & Lercker,
 d
t0 ¼ pffiffiffi G. (2003). Journal of Agricultural and Food Chemistry (in press).
SD= n Bonoli, M., Gallina Toschi, T., & Lercker, G. (2002). HPCE as rapid
and reliable tool for the determination of food natural antioxidant.
In Proceedings of the 15th International Symposium on Microscale
  Separation and Analysis-HPCE (p. 328). The Swedish Chemical
where d is defined as the average difference between Society Analytical Division.
total amount of catechins found by HPLC and that Corradini, C., & Cavazza, A. (1998). Applicazioni dell’elettroforesi
found by MEKC, SD is the standard deviation and n is capillare zonale (CZE) e della cromatografia elettrocinetica micel-
the number of replications. Finally, value was compared lare (MEKC) all’analisi degli alimenti. Italian Journal of Food
with t-Student value (t ¼0:05 Science, 299–316.
n1¼3 ¼ 2:35); the repeatability Dalluge, J. J., & Nelson, B. C. (2000). Determination of tea catechins.
hypothesis was rejected if t0 > t ¼0:05
n1¼3 . In fact, the sig- Journal of Chromatography A, 881, 411–424.
nificant difference between the total amount of catechins Faller, T., & Engelhardt, H. (1999). How to achieve higher repeat-
found with the two analytical methods can be attributed ability and reproducibility in capillary electrophoresis. Journal of
to the different peak area integration software used Chromatography A, 853, 83–94.
Gallina Toschi, T., Bordoni, A., Hrelia, S., Bendini, A., Lercker, G., &
(Faller & Engelhardt, 1999; Reijenga & Lee, 2001), and
Biagi, P. L. (2000). The protective role of different green tea extracts
to the higher efficiency and sensitivity of HPCE than after oxidative-damage is related to their catechin composition.
HPLC. Nevertheless, another factor that could have Journal of Agricultural and Food Chemistry, 48, 3973.
influenced these results could be the response variation Hertog, M. G. L., Freskens, E. J. M., Hollman, P. C. H., Katan,
related to the different wavelength used for the compo- M. B., & Kromhout, D. (1993). Dietary antioxidant flavonoids and
nent detection (200 and 270 nm for CE and HPLC, risk of coronary heart disease: the Zutphen elderly study. Lancet,
342, 1007.
respectively). Indeed, significant difference between the Hollman, P. C., Hertog, M. G., & Katan, M. B. (1996). Role of diet-
quantification of some compounds by the HPCE and ary flavonoids in protection against cancer and coronary heart dis-
HPLC methods could be due to interferences in the ease. Biochemical Society Transaction, 24, 785–789.
sample which could have a different weight at the two Horie, H., & Kohata, K. (2000). Analysis of tea components by
selected wavelength. As earlier mentioned, it was not high-performance liquid chromatography and high-performance
capillary electrophoresis. Journal of Chromatography A, 881, 425–
possible to use 200 nm as detection wavelength for both 438.
methods because the HPLC underwent interference by Issaq, H. J. (1999). Capillary Electrophoresis of Natural Products—II.
the mobile phase. Electrophoresis, 20, 3190–3202.
638 M. Bonoli et al. / Food Chemistry 81 (2003) 631–638

Larger, P. J., Jones, A. D., & Dacombe, C. (1998). Separation of tea Robards, K., & Antolovich, M. (1997). Analytical chemistry of fruit
polyphenols using micellar electrokinetic chromatography with bioflavonoids. Analyst, 122, 11R–34R.
diode array detection. Journal of Chromatography A, 799, 309– Sano, M., Tabata, M., Suzuki, M., Degawa, M., Miyase, T., &
320. Maeda-Yamamoto, M. (2001). Simultaneous determination of
Lee, B. L., & Ong, C. N. (2000). Comparative analysis of tea catechins twelve tea catechins by high-performance liquid chromatography
and theaflavins by high-performance liquid chromatography and with electrochemical detection. Analyst, 126, 816–820.
capillary electrophoresis. Journal of Chromatography A, 881, 439– Sesso, H. D., Gaziano, J. M., Buring, J. E., & Hennekens, C. H.
447. (1999). Coffee and tea intake and the risk of myocardial infarction.
Lindeberg, J. (1996a). Capillary electrophoresis in food analysis. Food American Journal of Epidemiology, 149, 162–167.
Chemistry, 55, 73–94. Stach, D., & Schmitz, O. J. (2001). Decrease in concentration of
Lindeberg, J. (1996b). Addendum to capillary electrophoresis in food free catechins in tea over time determined by micellar elec-
analysis. Food Chemistry, 55, 95–101. trokinetic chromatography. Journal of Chromatography A, 924,
Miura, Y., Chiba, T., Tomita, I., Koizumi, H., Miura, S., Umegaki, 519–522.
K., Hara, Y., Ikeda, M., & Tomita, T. (2001). Tea catechins prevent Wang, H., Helliwell, K., & You, X. (2000). Isocratic elution system for
the development of atherosclerosis in Apoprotein-E-deficient mice. the determination of catechins, caffeine and gallic acid in green tea
Journal of Nutrition, 131, 27. using HPLC. Food Chemistry, 68, 115–121.
Pelillo, M., Biguzzi, B., Bendini, A., Gallina Toschi, T., Vanzini, M., Worth, C. C. T., Wießler, M., & Schmitz, O. J. (2000). Analysis of
& Lercker, G. (2002). Preliminary investigation into development catechins and caffeine in tea extracts by capillary electrokinetic
with HPLC and MS-electrospray detection for the analysis of tea chromatography. Electrophoresis, 21, 3634–3638.
catechins. Food Chemistry, 78, 369–374. Zuo, Y., Chen, H., & Deng, Y. (2002). Simultaneous determination of
Reijenga, J., & Lee, H. K. (2001). Software and internet resources for catechins, caffeine and gallic acids in green, Oolong, black and pu-
capillary electrophoresis and micellar electrokinetic capillary chro- erh teas using HPLC with a photodiode array detector. Talanta, 57,
matography. Journal of Chromatography A, 916, 25–30. 307–316.

Você também pode gostar