Talanta 131 (2015) 624–631

Contents lists available at ScienceDirect

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

Sensitive analysis of amino acids and vitamin B3 in functional

drinks via field-amplified stacking with reversed-field stacking
in microchip electrophoresis
Minglei Wu a, Fan Gao a, Yi Zhang a, Qingjiang Wang a,n, Hui Li b,nn
a
Department of Chemistry, East China Normal University, 500 Dongchuan Road, Shanghai 200241, PR China
b
College of Chemistry and Chemical Engineering, Shanghai Jiaotong University, 800 Dongchuan Road, Shanghai 200240, PR China

art ic l e i nf o a b s t r a c t

Article history: An on-line preconcentration strategy combining field-amplified stacking and reversed-field stacking was
Received 27 June 2014 developed for efficient and sensitive analysis of amino acids and vitamin B3 including lysine (Lys),
Received in revised form taurine (Tau), and niacinamide (NA) by microchip electrophoresis with LIF detection. In this technique,
16 August 2014
the addition of a reversed-polarity step termed reversed-field stacking could enhance the preconcen-
Accepted 18 August 2014
tration effect of field-amplified stacking and push most of the sample matrix out of the separation
Available online 23 August 2014
channel, thus greatly improving the sensitivity enhancement by 1–2 orders of magnitude over the
Keywords: classical MCE–LIF methods. The related mechanism as well as important parameters governing
Field-amplified stacking preconcentration and separation have been investigated in order to obtain strongest sensitivity
Reversed-field stacking amplification and maximum resolution. Under optimal conditions, all analytes were successfully focused
Amino acids
and completely separated within 4 min. The limits of detection for Lys, Tau, and NA were 0.25, 0.50, and
Vitamin B3
0.20 nM (S/N¼ 3), respectively, and enhancement factors of 165-, 285-, and 236-fold were obtained for
Functional drinks
Lys, Tau, and NA as compared to using the no concentration step. Other validation parameters such as
linearity and precision were considered as satisfactory. The proposed method also gave accurate and
reliable results in the analysis of these functional ingredients in eight functional drink samples.
& 2014 Elsevier B.V. All rights reserved.

1. Introduction development of pellagra [3,4]. Lys is a building block for proteins

Functional drinks are non-alcoholic drinks fortified with func-
tional ingredients, which are promoted with benefits such as heart
health, improved immunity and digestion, joint health, satiety, and
energy boosting [1,2]. Nowadays, the consumers' demand for
functional drinks is clearly rising [2] and the fortification of
functional drinks with vitamins and amino acids is very common
in the modern food industry. Niacinamide (vitamin B3), which
exists in various kinds of functional drinks, is an important
endogenous metabolite and dietary component. Niacinamide
(NA) plays an important role in the cellular energy metabolism
and influences oxidative stress. In deficiency states, inadequate NA
leads to fatigue, loss of appetite, pigmented rashes of the skin, and
oral ulcerations [3]. More severe states of deficiency cause the
Abbreviations: Cy5, sulfoindocyanine succinimidyl ester; FASS, field-amplified
stacking; Lys, lysine; MCE, microchip electrophoresis; NA, niacinamide; Tau, taurine
n
Corresponding author. Tel.: þ 86 21 54340015.
nn
Corresponding author. Tel.: þ 86 21 54743271.
E-mail addresses: qjwang@chem.ecnu.edu.cn (Q. Wang),
chemlihui@sjtu.ecnu.edu.cn (H. Li).
and an essential amino acid that cannot be synthesized by the
human body, making it necessary to be supplied by external
sources. For children, Lys is essential for healthy growth and
development especially for proper growth of bones. In addition,
it helps the body absorb calcium, maintain the proper nitrogen
balance, and ensure lean body mass [5,6]. As a member of the
family of sulfur-containing amino acids, Tau is naturally present in
abundant concentrations in mammalian tissues [7,8]. This com-
pound has been used as an additive in functional drinks because of
its many important physiological functions, including the forma-
tion of bile salts, modulation of calcium flux, and neuronal
excitability, osmoregulation, and immune function [9]. Some
therapeutic effects have also been proposed for Tau supplementa-
tion such as treatment for diabetes [10], high blood pressure [11],
heart failure [12], retinal degeneration [13], and skeletal muscle
disorders [14].
The need for simultaneous reliable and sensitive analytical
methods for the health promoting compounds in functional drinks
is obvious. HPLC is the most common technique for separating
micronutrients or amino acids in food products, which has been
used for separating vitamins in functional drinks [15] and amino

http://dx.doi.org/10.1016/j.talanta.2014.08.051
0039-9140/& 2014 Elsevier B.V. All rights reserved.
 M. Wu et al. / Talanta 131 (2015) 624–631
acids in food, beverages and feedstuff [16,17]. However, this
method usually requires larger sample volume with much more
solvent consumption. In recent decades, CE has been accepted as a
powerful analytical technique in analyzing nutritious compounds
due to its lower consumption of sample and better separation
efficiency compared to HPLC [18]. Szymom et al. presented a CE-
based method for separation and sensitive analysis of vitamins B
[19]. Additionally, Zhao et al. demonstrated separation and deter-
mination of vitamins B and essential amino acids in health drinks
using CE [20].
In recent years, the development of micro-fluidic devices, also
called “lab-on-a-chip”, has witnessed an explosive growth since
their introduction. Microchip electrophoresis (MCE), which is one
of the most important applications of micro-fluidics, is increas-
ingly being viewed as a new trend in food analysis [21]. Microchips
typically consume picoliters of samples [22], which is much less
than the sample consumption of CE (at the nanoliter level) and
HPLC (at the microliter level). In addition, MCE offers significant
benefits such as miniaturization, reduced reagent consumption,
and fast operation. Regarding the detection system, LIF has proved
to be a potent detection method in terms of its high sensitivity and
high selectivity in real sample analysis. However, the sensitivity of
MCE–LIF is sometimes not sufficient because of the characteristi-
cally low sample volume and short optical pathlength of the
microchannels [23]. Therefore, various on-line preconcentration
methods have been developed and implemented to enhance the
sensitivity of MCE [24,25], including field-amplified stacking
(FASS) [26], transient isotachophoresis [27], isoelectric focusing
[28], transient trapping [29], sweeping [30], reversed-field stack-
ing [31], and so on. Such techniques can be performed individually
or in combination, affording ten- to almost one million-fold
improvements in detection sensitivity. Qiong et al. reported a
highly effective on-chip preconcentration method combining FASI
and bovine serum albumin sweeping for ultrasensitive detection
of green fluorescent protein on a simple cross-channel microchip
device, which yielded a concentration factor of 3570 [30]. Our
group recently reported a multiple-concentration method com-
bining chitosan sweeping, reversed-field stacking, and FASS for
ultrasensitive detection of bacteria by MCE–LIF, which achieved
nearly 6000-fold signal enhancement [31].
The applications of FASS to CE have been well described to
provide high sensitivity enhancement in a number of studies
[32,33]. However, microchip-based separation systems have short
separation distances that are 10  100 times shorter than those of
standard capillary-based systems, limiting the concentration
enhancement of FASS. In addition, the preconcentration effect of
FASS is detrimental to the separation efficiency of MCE, leading to
O O
H 2N
OH
NH2
Lysine
H3C CH3 H3C CH
3
N N
O3S CH3
O O
Cy5 NHS ester O O N
O
Fig. 1. Chemical structures of the analytes assayed and scheme of the labeling reaction with Cy5.
625
the decline in resolution. In this case, the addition of a reversed-
polarity step called reversed-field stacking could lengthen the
effective separation distance and provide a longer time for FASS by
pushing back the sample plug, thus enhancing the enrichment
effect of FASS and improving the resolution of all analytes.
Simultaneously, most of the low-conductivity sample matrix was
pumped out of the separation channel by using reversed-field
stacking, thus further improving the concentration enhancement.
Herein, we applied an on-line preconcentration approach
combining FASS and reversed-field stacking to the MCE–LIF
analysis of three health promoting compounds, including Lys,
Tau, and NA (Fig. 1). To the best of our knowledge, this is the first
report of combination of the FASS-reversed-field stacking techni-
que for the simultaneous determination of varied compounds in
food samples by MCE. The sensitivity obtained by this strategy was
improved by one to two orders of magnitude as compared to those
of previously reported MCE–LIF methods [34,35]. Under optimal
conditions, the two amino acids and vitamin B3 were successfully
focused and well separated within 4 min, and the detection
sensitivities were improved 165-, 285-, and 236-fold for Lys, Tau,
and NA using this concentration technique as compared to using a
no concentration method. The proposed method allows fast and
sensitive analysis of these amino acids as well as vitamin B3 and it
was successfully applied to the determination of the three func-
tional ingredients in eight functional drinks. Therefore, it is self-
evident that this approach could potentially be used for the
simultaneous analysis of low-abundance beneficial compounds
in food products.
2. Materials and methods
2.1. Apparatus and chemicals
All MCE experiments were conducted using a homemade MCE
system coupled with a laser-induced fluorescence (LIF) detection
device that was created by Shanghai Spectrum Ltd. Co., Zhejiang
University and our research group [36]. Briefly, a diode laser
(5 mW) was used to generate an excitation beam at 635 nm. The
fluorescence signal was spectrally isolated using an edge filter and
was subsequently collimated with an achromatic lens before being
focused onto the photomultiplier tube (PMT). The high voltage
power unit variable in the range 0 76 kV was used for on-chip
sample injection and zone electrophoretic separation. The ampli-
fied current was transferred directly through a 10 kΩ resistor to a
24 bit A/D interface at 10 Hz (Borwin, JMBS Developments, Le
Fontanil, France) and stored in a personal computer.
O
OH
S N NH2
H 2N O
Taurine Niacinamide
O O
SO3
+ R-NH2 Cy5 NH-R + HO N
626 M. Wu et al. / Talanta 131 (2015) 624–631
Analytical-grade chemicals were used unless otherwise stated.
Lysine (Lys), taurine (Tau), and niacinamide (NA) were purchased
from Sigma (St. Louis, MO, USA). Sodium tetraborate (Na2B4O7),
sodium hydroxide (NaOH), hydrochloric acid, and acetonitrile were
from Sinopharm Chemical Reagent Co. Ltd. (Shanghai, China). Sul-
foindocyanine succinimidyl ester (Cy5) was obtained from GE
Healthcare Company (Shanghai, China). All solutions were prepared
by ultrapure water supplied by Milli-Q water purification system
(Millipore, Bedford, MA, USA).
2.2. Solutions
A running buffer solution was prepared by dilution of the
appropriate amount of Na2B4O7 in ultrapure water to a concentration
of 100 mM, whereas sample buffer and derivatization buffer solu-
tions were made by dissolving Na2B4O7 with water to 10 mM and
20 mM, respectively. Desired pH values of these buffer solutions
were obtained with 1 M NaOH or 0.5 M hydrochloric acid. Standard
solutions of individual Lys, Tau, and NA were prepared with a
concentration of 100 mM in ultrapure water and diluted to 100 μM
with derivatization buffer. The stock solution Cy5 (100 μM) was
made by dissolving in anhydrous acetonitrile. The above solutions
were sealed and stored at 4 1C in a refrigerator.
2.3. Procedures of derivatization
Standard solutions of 100 mM Lys, Tau, and NA were diluted to
100 μM with derivatization buffer (20 mM borate solution at pH
8.60). Derivatization of individual compounds was performed by
mixing diluted solution with 100 μM Cy5 solution 1:1 (v:v) in a
1.5 mL centrifuge tube. The three mixtures reacted in darkness for
4 h at room temperature, and then diluted to required concentra-
tions with sample buffer (10 mM borate solution at pH 9.88)
before using.
2.4. Preparation of drink sample
Eight functional drink samples purchased from supermarkets in
Shanghai were named F1, F2, F3, F4, F5, F6, F7, and F8, respectively.
The functional drink samples were diluted two-fold with 40 mM
borate solution and adjusted to pH 8.60 with 1 M NaOH. Finally, the
diluted sample solutions were derivatized as stated in Section 2.3,
and then diluted to a certain concentration with sample buffer
(10 mM borate solution at pH 9.88) before analysis.
1.30kv 0.80kv
1 1
L1 L1
1.30kv 3 2 1.80kv 0.50kv 3 2 0.50kv
L3 L2 L3 L2
L4 L4
4 4
Ground Ground
Fig. 2. Schematic diagram of the sample loading, on-line preconcentration, and separation of nutritious compounds: (A) preloading, (B) FASS, (C) reversed-field stacking, and
(D) separation. The dark blue zone represents the concentrated sample by FASS before using reversed-field stacking technique, the black zone represents the area of the
concentrated sample after reversed-field stacking, the light blue zone represents the sample matrix, and the clear zone represents the running buffer. 1 ¼ sample reservoir;
2¼ buffer reservoir; 3 ¼sample waste reservoir; 4¼ buffer waste reservoir. L 1–4 represent four channels of microchips. Arrows indicate bulk flow directions. The voltages
(kV) applied to the reservoirs at each step are indicated. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this
article.)
2.5. MCE conditions
The glass microchip design used in these experiments con-
sisted of a simple cross-channel. The separation channel was
60 mm in length and 45 mm from the injection intersection to
the detection point. All other channels had a length of 10 mm
measured from the channel intersection. Microchannels were
etched to a depth of 25 μm and a width of 70 μm. Platinum
electrodes were inserted into the reservoirs, providing electrical
contact from the power supply to the electrolyte solutions. All
experiments were running in full filling mode.
Before a new microchip was used for the first time, it was
washed with 98% H2SO4 and ultrapure water for 10 min. Subse-
quently, the channels were flushed with 1 M NaOH for 20 min,
ultrapure water for 10 min, and the MCE running buffer for 10 min.
Before each injection, the microchannels were washed for 2 min
with 1 M NaOH, for 1 min with ultrapure water and for 2 min with
MCE running buffer.
2.6. Procedure of on-line preconcentration method
For FASS, the samples were prepared with an appropriate
sample buffer (10 mM borate solution) with a much lower con-
ductivity compared with running buffer (100 mM borate solution).
The reversed-field stacking step was maintained similar to the
sample loading time, so that the sample plug could move to the
injection cross within an appropriate distance. A four-step MCE
procedure with the voltage configuration of each step was
designed as shown in Fig. 2, which was responsible for the
preconcentration effect in the MCE method.
3. Results and discussion
3.1. Field-amplified stacking-reversed-field stacking mechanism
In order to realize the highly efficient detection of three beneficial
compounds, an on-line multiple-preconcentration method combining
FASS and reversed-field stacking was employed. The schematic
mechanism of this method is shown in Fig. 2. At the beginning, the
voltages of each of the reservoirs were adjusted to ensure that the
low-conductivity sample solution flowed from the sample reservoir to
the sample waste reservoir through the injection cross and the
running buffer solution flowed from the buffer reservoir to both waste
reservoirs (Fig. 2A). The voltage configuration was then changed to the
0.20kv 1.30kv
1 1
L1 L1
Ground 3 2 0.20kv 1.30kv 3 2 1.80kv
L3 L2 L3 L2
L4 L4
4 4
0.80kv Ground
 M. Wu et al. / Talanta 131 (2015) 624–631
loading step. During the loading step, as shown in Fig. 2B, a gated
injection method was selected to introduce the sample plug into the
separation channel. Meanwhile, FASS occurred in the separation
channel (the dark blue zone in Fig. 2B). In the process of FASS, the
samples were prepared in a medium of lower conductivity than the
running buffer. Charged analytes move more quickly in the low-
conductivity sample zone and are immediately retarded at the inter-
face between the two buffers with different ionic strengths and
electric fields. However, the micro-fluidic devices have short separa-
tion distances, limiting the concentration enhancement of FASS. In
addition, the enrichment effect of FASS is detrimental to the separation
efficiency of MCE, leading to the decline in separation resolution of the
analytes. To solve these problems, an additional step which was
termed reversed-field stacking was performed by switching the
polarity (Fig. 2C). Under the reversed-field stacking conditions, the
concentrated sample zone in L4 was pushed back to the injection cross
within an appropriate distance, which could give analytes a longer
time for boundary enrichment and lengthen the effective separation
distance, thus enhancing the enrichment effect of FASS and improving
separation efficiency. Simultaneously, most of the long and vacant
sample matrix was pumped from the separation channel (L4) into L3
in the process of reversed-field stacking, thus avoiding the reduction of
effective separation length and leading to further concentration
enhancement.
These two steps of the on-line concentration method were
crucial in order to obtain the highest concentration enhancement.
Fig. 3 displays the electrophoretic profiles for the different con-
centration steps. The standard electropherograms shown in Fig. 3B
and C were obtained by 10-fold dilution of the sample used in
Fig. 3A. The electropherogram of the typical gated injection
method without any concentration is shown in Fig. 3A. For FASS
(Fig. 3B), all analytes were enriched and the signal intensities
increased. However, the decline in separation resolution can be
observed in the magnifications attached to Fig. 3A and B. For the
combined FASS and reversed-field stacking (Fig. 3C), signals of all
analytes were greatly enhanced and baseline separation was
realized for all analytes. These observations demonstrated the
high efficiency of this multiple-concentration method.
Fig. 3. Signal enhancement of the multiple concentration: (A) signal intensity
without concentration. The running buffer was 100 mM borate solution at pH 9.88
and the sample was diluted with the running buffer solution. The sample injection
time was 2 s. (B) Signal intensity with FASS. The sample was prepared in a 10-fold-
diluted running buffer, and the sample injection time was 2 s. (C) Signal intensity
with a combination of FASS and reversed-field stacking. The sample injection time
was 10 s, and the reversed-polarity time was 8 s. The concentrations of Lys, Tau,
and NA in (A) were 0.6, 0.9, and 0.6 μM, respectively. The sample concentrations in
(B) and (C) were 1/10 of that in (A). Peak identification: 1, the excess of Cy5; 2, Lys;
3, Tau; and 4, NA.
627
3.2. Optimization of derivatization conditions
As the analytes have no native fluorescence, they were deriva-
tized with sulfoindocyanine succinimidyl ester (Cy5). The scheme
of the labeling reaction with Cy5 can be seen in Fig. 1. The
derivatization reaction was optimized with respect to the deriva-
tization borate buffer, time, and temperature to enhance the
detection sensitivity. The composition of the derivatization borate
buffer was optimized by varying the pH and the concentration of
borate solution over the range of 7.5–9.5 and 10–50 mM. It was
found that the derivatization efficiency increased with increasing
pH values up to 8.6 and then decreased up to a pH value of 9.5,
which can be attributed to the hydrolytic degradation of the Cy5
[37]. Therefore, pH 8.6 was selected as optimum. From the
experiments, 20 mM derivatization borate buffer was found to
achieve maximum reaction efficiency in the labeling reaction.
Further increases in the concentration of derivatization buffer
provided a decrease in the reaction yield. Time and temperature
played significant roles in the labeling process. It was found that
the maximum yield was obtained at room temperature by apply-
ing 4 h of reaction time. These conditions were used for all
subsequent experiments.
3.3. Optimization of MCE separation conditions
3.3.1. Effect of running buffer concentration
A series of running buffer consisting of 25, 50, 75, 100 mM
borate solution (pH 9.26) was prepared to study the effect of
running buffer concentration on the separation of Lys, Tau, and NA.
Higher concentrations of running buffer were not considered
because higher concentration buffer resulted in higher current,
which generated the high Joule heat causing the instability of
baseline. As shown in Fig. S1 (see Supplementary material),
resolution was improved with the increase in buffer concentration.
The peaks of Lys and the excess of Cy5 were separated with the
higher concentrations and the best resolution of Tau and NA was
achieved when the buffer concentration was increased to 100 mM
(Fig. S1D). As a result, 100 mM borate solution was selected as
optimum and used for the following experiments.
3.3.2. Effect of running buffer pH value
In MCE, the pH value of running buffer is an important
parameter to be optimized as it affects both the charge of the
analytes and EOF, which in turn plays a critical role in electro-
phoretic mobility and the separation of analytes. For the separa-
tion of Lys, Tau, NA, and the excess of Cy5, the effect of buffer pH
values ranging from 9.26to10.85 was studied. It can be seen from
Fig. 4 that the baseline separation of all analytes was achieved
using 100 mM borate buffer at pH 9.88 (Fig. 4D). At pH lower than
9.8, the labeled Tau and NA co-migrated and overlapped, while at
pH equal to or higher than 10, the baseline was unstable and the
resolution was poor because of the high Joule heating. It was
obvious that the perfect separation was obtained at pH 9.88, which
was determined to be the best pH for MCE separation.
3.4. Optimization of on-line preconcentration conditions
3.4.1. Effect of sample buffer concentration
Theoretically, the concentration enhancement depends on the
concentration ratio of running buffer to sample buffer, and higher
ratios result in stronger fluorescent signals of analytes in FASS [38].
However, in a realistic system, the increase in concentration ratio of
running buffer to sample buffer should be balanced against the
negative effect of dispersion and advection on stacking efficiency
caused by the mismatch of electroosmotic velocity [39]. Therefore,
628 M. Wu et al. / Talanta 131 (2015) 624–631

Fig. 4. Effect of pH of the running buffer on the separation of three Cy5-labeled nutritious compounds. The pH values of running buffer solution: (A) 9.26, (B) 9.42, (C) 9.65,
and (D) 9.88. Buffer concentration: 100 mM. Concentration of Lys, Tau, and NA were 0.03, 0.03, and 0.025 μM, respectively. The samples were prepared in a 10-fold-diluted
running buffer. Other conditions are the same as those in Fig. 3C.

under the optimized separation conditions (100 mM borate solution

at pH 9.88 in running buffer), sample buffer concentration has been
studied. Different concentrations of Na2B4O7 (ranging from 10 to
75 mM) in sample buffer were investigated by testing the fluorescent
signal intensity of NA. From Fig. 5A–D, it can be observed that the
peak height increased with the decrease in sample buffer concentra-
tion and the best result for signal amplification was obtained with
10 mM borate solution as the sample buffer (Fig. 5D). Further
decreases in sample buffer concentration did not serve to improve
the enrichment effect in our experiments. Thus, the optimal sample
buffer was chosen as 10 mM borate solution (pH 9.88).

3.4.2. Effect of injection time and reversed-polarity time

Injection time and reversed-polarity time have great influence on
signal amplification (Fig. 6). As mentioned previously, both the
injection step and reversed polarity step were accompanied by FASS,
and the reversed polarity step using reversed-field stacking techni-
que could provide stronger signal amplification by enhancing the Fig. 5. Effect of the sample buffer concentration on signal enhancement. The
concentration of borate buffer solution in sample buffer: (A) 75 mM, (B) 50 mM,
enrichment effect of FASS and pushing most of the sample matrix out (C) 20 mM, (D) 10 mM. Concentration of NA was 0.07 μM. The samples were diluted
of the separation channel. However, with a longer reversed-polarity with the sample buffer solution and all buffer pH values were 9.88. Other
time, the analytes would potentially be pumped from the separation conditions are the same as those in Fig. 3C.
 M. Wu et al. / Talanta 131 (2015) 624–631 629

channel with the sample matrix, leading to the reduction of all peak broadening of signals with a longer injection time, the optimal
heights. After comprehensively considering every factor, the selected injection time and reversed-polarity time were selected to be 10 s
reversed-polarity time was 2 s shorter than the injection time, which and 8 s, respectively (Fig. 6C).
made the sample plug move to the injection cross within an
appropriate distance to avoid pushing out the concentrated sample. 3.5. Method validation
As shown in Fig. 6, the increase in injection time and reversed-
polarity time could strengthen the signal intensities of all analytes. The performance and reliability of the proposed method were
Considering the decrease in separation efficiency due to band assessed by determining linearity, limit of detection (LOD), preci-
sion (RSD) for amino acids and vitamin B3. Table 1 gives the
equations for the calibration curves obtained by plotting the peak
height against analyte concentration in the linear range, the LODs
calculated as the minimum analyte concentration providing sig-
nals 3 times the background noise, the intraday precision obtained
by repeating the analysis three times, and the sensitivity enhance-
ment factors of all analytes using the multiple-concentration
method (Fig. 3C) compared to, when no concentration was
performed (Fig. 3A). Under optimized conditions, the method
proposed allowed the determination of these compounds at very
low levels (LODs ranged from 0.20 to 0.50 nM) and the linearity
was satisfactorily studied with correlation coefficients from 0.9958
to 0.9988. The RSDs of the migration time and peak height were in
the range of 1.2–3.4% and 1.3–2.7%, respectively, indicating good
reproducibility and high precision. The sensitivity enhancement
factors in peak heights for Lys, Tau, and NA using this concentra-
tion method were 165, 285, and 236, respectively.

3.6. Method application for the analysis of functional drink samples

To demonstrate how the proposed MCE–LIF method can be

Fig. 6. Effect of injection time and reversed-polarity time on peak intensity.
Injection time and reversed-polarity time: (A) 4 s, 2 s; (B) 8 s, 6 s; and (C) 10 s,
applied for the analysis of the real sample, the method was further
8 s. Concentration of Lys, Tau, and NA were 0.015, 0.025, and 0.015 μM, respectively. investigated by the analysis of functional ingredients in eight
Other conditions are the same as those in Fig. 3C. functional drink samples. The samples were prepared according

Table 1
Analytical performance and enrichment factor results of the multiple-concentration MCE method.

Compound Linearity range (μM) Linear regression a(n¼ 10) Correlation coefficients (R2) LOD (nM) RSD% (n¼ 3)b MT/peak height Enrichment factor

Lys 0.003–0.10 y¼28307x  8.4514 0.9958 0.25 3.4/1.3 165

Tau 0.004–0.15 y¼14678x  33.3098 0.9988 0.50 2.0/0.8 285
NA 0.003–0.10 y¼40174x þ5.0533 0.9982 0.20 1.2/2.7 236

a
In the regression equation, the x value was the concentration of analytes (μM), the y value was the peak height (mv).
b
MT, migration time. Concentrations of Lys, Tau, and NA were 0.013, 0.020, and 0.013 μM, respectively. Other conditions were the same as those in Fig. 3C.

3500 3500 3500 3500

1
3000 3000 3000 3000 1

2500 2500 1 2500 2500

2000 3 2000 2000 1 2000

4
1500 1500 4 1500 1500
2 4
1000 1000 1000 1000
Fluorescence Intensity (mv)

4
500 500 500 500

0 0 0 0

0 50 100 150 200 0 50 100 150 200 0 50 100 150 200 0 50 100 150 200

3500 3500 3500 3500

3000 3000 3000 3000

2500 2500 2500 2500

2000 2000 2000 2000

1500 1500 1500 1500 4

3 1
3
1000 1 1000 1 1000 1 1000
3
500 500 500 500
4 4 2 4
0 0 0 0

0 50 100 150 200 0 50 100 150 200 0 50 100 150 200 0 50 100 150 200

Time (s)

Fig. 7. Electropherograms of eight Cy5-labeled functional drink samples after dilution. (A) F1, (B) F2, (C) F3, (D) F4, (E) F5, (F) F6, (G) F7, and (H) F8. Dilution folds of eight
drink samples are listed in Table 2. Other experiment conditions are the same as those in Fig. 3C.
630 M. Wu et al. / Talanta 131 (2015) 624–631

Table 2 two amino acids and vitamin B3 in functional drinks. Using this
Quantitative determination of beneficial compounds in eight functional drinks. method, three functional ingredients could be simultaneously focused
on and separated within 4 min. Under optimal conditions, the strategy
Sample Functional Concentrationa Amount Amount Recovery
ingredient detected (μM) added found (%) allowed for the determination of these beneficial compounds at very
after dilution (μM) (μM) low concentrations (0.20–0.50 nM) and afforded approximately 150-
to 300-fold improvements in peak height. Moreover, the method was
F1 Lys 0.0424 0.0550 0.0978 100.7 successfully applied to functional drink samples with a satisfactory
Tau 0.1201 0.0250 0.1454 101.2
NA 0.0105 0.0095 0.0198 97.9
recovery. Compared with normal detection methods for beneficial
F2 NA 0.0350 0.0350 0.0713 103.7 compounds in food analysis, this multiple-concentration MCE method
F3 NA 0.0285 0.0200 0.0479 97.0 has many merits, including miniaturization, rapid operation, low
F4 NA 0.0380 0.0250 0.0642 104.8 sample consumption, and high sensitivity.
F5 Tau 0.0819 0.0400 0.1249 107.5
NA 0.0039 0.0125 0.0161 97.6
F6 Tau 0.0335 0.0300 0.0645 103.3
NA 0.0046 0.0100 0.0145 99.0
F7 Lys 0.0056 0.0100 0.0159 103.0 Acknowledgments
Tau 0.0758 0.0500 0.1252 98.8
NA 0.0030 0.0025 0.0056 104.0
F8 NA 0.0343 0.0200 0.0558 107.5
This work was financially supported by the Program New
Century Excellent Talents in University (Grant NCET-08-0191)
a
The concentrations of the functional ingredients in eight functional drink and the National Program on Development of Scientific Instru-
samples were determined after dilution. Dilution fold: F1, 32000; F2, 1600; F3, ments and Equipment (Grant 2011YQ150072).
2000; F4, 2000; F5, 32000; F6, 64000; F7, 320000; and F8, 2000.

Appendix A. Supplementary matrials

to procedures in Section 2.4 and analyzed by using the multiple-
concentration method under optimized conditions. All drink Supplementary data associated with this article can be found in
samples were diluted before analysis in order to avoid instability the online version at http://dx.doi.org/10.1016/j.talanta.2014.08.
of baseline and interference with resolution caused by impurities 051.
in real samples. Fig. 7 shows the electropherograms obtained from
analysis of the eight functional drinks. Because the migration time References
of the analytes varied in the real sample analysis, the peaks were
verified by a standard addition method. For F1 and F7, all three [1]
functional ingredients (Lys, Tau, and NA) were detected. Both Tau [2]
[3]
and NA were determined in F5 and F6, while only NA was detected [4]
in sample F2, F3, F4, and F8. The components detected in eight
drink samples were consistent with the components listed in the [5]
labels of nutrition fact. [6]
Concentrations of three beneficial compounds in eight func-
tional drink samples and recoveries are summarized in Table 2. [7]
Satisfactory recoveries (97.0–107.5%) were obtained in real drink [8]
samples, indicating that real sample matrices had virtually no [9]
effect on the performance of the method proposed. Besides, the
[10]
concentrations of functional ingredients detected by the proposed [11]
method met the standard of the values claimed in nutrition facts
of all drink samples except for F6. In F6, the Tau content detected [12]
was about half of the labeled value. The concentrations of each
[13]
functional ingredient in the drink samples were determined with
the calibration equations elaborated in Section 3.5 (Table 1). Since
little has been reported on the simultaneous separation and [14]
determination of these beneficial compounds in functional drinks, [15]
the proposed MCE method was compared with HPLC and CE by
comparing the analysis of NA found in the same segment of [16]
[17]
functional beverages like energy drinks [20,40]. The NA content [18]
detected by the method developed was similar to the assay results
found by HPLC and CE, and the analysis time of NA assayed in real [19]
[20]
drink samples using the proposed MCE method was within 3 min, [21]
which was less than that using HPLC (about 4 min) and CE (about
22 min) [20,40]. In summary, these results testify to the good [22]
[23]
performance of the proposed method in the determination of [24]
health promoting compounds in this type of sample. [25]
[26]
[27]
[28]
4. Conclusions [29]
[30]
[31]
We presented an on-line preconcentration MCE method combin-
ing FASS and reversed-field stacking for the simultaneous analysis of [32]
I. Siro, E. Kapolna, B. Kapolna, A. Lugasi, Appetite 51 (2008) 456–467.
J. Wan, Food Ferment. Ind. 33 (2007) 101–106.
K. Maiese, Z.Z. Chong, J. Hou, Y.C. Shang, Molecules 14 (2009) 3446–3485.
L.F. Sallum, F.L. Felipe Soares, J.A. Ardila, R.L. Carneiro, Talanta 118 (2014)
353–358.
S. Harakeh, M. Diab-Assaf, K. Abu-El-Ardat, A. Niedzwiecki, M. Rath, Chem.
Biol. Interact. 164 (2006) 102–114.
M. Kumar, M. Ali Khan, Z.Al. Othman, J. Chem. Technol. Biotechnol. 88 (2013)
910–918.
M. Hohmann, C. Felbinger, N. Christoph, H. Wachter, J. Wiest, U. Holzgrabe, J.
Pharm. Biomed. Anal. 93 (2014) 156–160.
T.R.I. Cataldi, G. Telesca, G. Bianco, D. Nardiello, Talanta 64 (2004) 626–630.
Y.-R. Yu, X.-C. Liu, J.-S. Zhang, C.-Y. Ji, Y.-F. Qi, Parasitol. Res. 112 (2013)
3457–3463.
S.J. Kim, C. Ramesh, H. Gupta, W. Lee, J. Biol. Reg. Homeos. Ag. 21 (2007) 63–77.
A. El Idrissi, E. Okeke, X. Yan, F. Sidime, L.S. Neuwirth, Adv. Exp. Med. Biol. 775
(2013) 407–425.
M.R. Beyranvand, M.K. Khalafi, V.D. Roshan, S. Choobineh, S.A. Parsa, M.
A. Piranfar, J. Cardiol. 57 (2011) 333–337.
N. Froger, L. Cadetti, H. Lorach, J. Martins, A.-P. Bemelmans, E. Dubus,
J. Degardin, D. Pain, V. Forster, L. Chicaud, I. Ivkovic, M. Simonutti, S. Fouquet,
F. Jammoul, T. Leveillard, R. Benosman, J.-A. Sahel, S. Picaud, Plos One 7 (2012).
D.C. Camerino, D. Tricarico, S. Pierno, J.F. Desaphy, A. Liantonio, M. Pusch, R. Burdi,
C. Camerino, B. Fraysse, A. De Luca, Neurochem. Res. 29 (2004) 135–142.
J.A. Mendiola, F.R. Marin, F.J. Senorans, G. Reglero, P.J. Martin, A. Cifuentes,
E. Ibanez, J. Chromatogr. A 1188 (2008) 234–241.
D. Heems, G. Luck, C. Fraudeau, E. Verette, J. Chromatogr. A 798 (1998) 9–17.
E. Barrado, J.A. Rodriguez, Y. Castrillejo, Talanta 78 (2009) 672–675.
M. Castro-Puyana, A.L. Crego, M.L. Marina, C. Garcia-Ruiz, Electrophoresis 28
(2007) 4031–4045.
S. Dziomba, P. Kowalski, T. Baczek, J. Chromatogr. A 1267 (2012) 224–230.
D. Zhao, M. Lu, Z. Cai, Electrophoresis 33 (2012) 2424–2432.
A. Escarpa, M.C. Gonzalez, A.G. Crevillen, A.J. Blasco, Electrophoresis 28 (2007)
1002–1011.
V. Dolník, S. Liu, S. Jovanovich, Electrophoresis 21 (2000) 41–54.
K. Sueyoshi, F. Kitagawa, K. Otsuka, Anal. Sci. 29 (2013) 133–138.
K. Sueyoshi, F. Kitagawa, K. Otsuka, J. Sep. Sci. 31 (2008) 2650–2666.
F. Kitagawa, T. Kawai, K. Sueyoshi, K. Otsuka, Anal. Sci. 28 (2012) 85–93.
B. Jung, R. Bharadwaj, J.G. Santiago, Electrophoresis 24 (2003) 3476–3483.
Z. Mala, P. Gebauer, P. Bocek, Electrophoresis 34 (2013) 19–28.
M. Kinoshita, Y. Nakatsuji, S. Suzuki, T. Hayakawa, K. Kakehi, J. Chromatogr. A
1309 (2013) 76–83.
K. Sueyoshi, F. Kitagawa, K. Otsuka, Anal. Chem. 80 (2008) 1255–1262.
Q. Pan, M. Zhao, S. Liu, Anal. Chem. 81 (2009) 5333–5341.
Z.-F. Wang, S. Cheng, S.-L. Ge, H. Wang, Q.-J. Wang, P.-G. He, Y.-Z. Fang, Anal.
Chem. 84 (2012) 1687–1694.
X. Zheng, M. Lu, L. Zhang, Y. Chi, L. Zheng, G. Chen, Talanta 76 (2008) 15–20.
 M. Wu et al. / Talanta 131 (2015) 624–631 631

[33] Y. He, X. Li, P. Tong, M. Lu, L. Zhang, G. Chen, Talanta 104 (2013) 97–102.
[34] Y. Huang, M. Shi, S. Zhao, H. Liang, J. Chromatogr. B 879 (2011) 3203–3207.
[35] M. Wu, S. Cheng, G. Li, Z. Wang, Q. Wang, P. He, Y. Fang, Chin. J. Chem. 30
(2012) 2517–2521.
[36] G. Li, S. Ge, X. Ni, S. Dong, Q. Wang, P. He, Y. Fang, Chin. J. Chem. 27 (2009)
2207–2211.
[37] E. Orejuela, M. Silva, Electrophoresis 26 (2005) 4478–4485.
[38] D.S. Burgi, R.L. Chien, Anal. Chem. 63 (1991) 2042–2047.
[39] R. Bharadwaj, J.G. Santiago, J. Fluid. Mech. 543 (2005) 57–92.
[40] A. Gliszczyńska-Świgło, I. Rybicka, Food Anal. Methods (2014), http://dx.doi.
org/10.1007/s12161-014-9880-0 (Published online: 8 May 2014).