CHINESE JOURNAL OF CATALYSIS

Volume 33, Issue 11, 2012

Online English edition of the Chinese language journal

Cite this article as: Chin. J. Catal., 2012, 33: 1783–1790. ARTICLE

A Carbon Nanotube Modified Electrode for Determination of

Caffeine by Differential Pulse Voltammetry
Biuck HABIBI1,*, Mehri ABAZARI1, Mohammad Hossien POURNAGHI-AZAR2
1
Electroanalytical Chemistry Laboratory, Department of Chemistry, Faculty of Sciences, Azarbaijan University of Tarbiat Moallem,
Tabriz 53714-161, Iran
2
Center of Excellence of New Materials and Clean Chemistry, University of Tabriz, Tabriz 5166616471, Iran

Abstract: A simple and sensitive electrochemical detection of caffeine (CAF) using a single-walled carbon nanotubes on carbon-ceramic
electrode (SWCNT/CCE) is reported. CAF was oxidized at the surface of the modified electrode to produce an anodic peak at 1.38 V versus
the saturated calomel electrode in 0.01 mol/L, pH 1.7 H2SO4 solution in cyclic voltammetry. The experimental parameters, namely, type of
electrolyte, pH value, and amount of SWCNTs casted, were optimized. Using the optimum conditions, the anodic peak current in differ-
ential pulse voltammetry was linear with CAF concentration in the range of 2.5 × 1071.0 × 104 mol/L. The detection limit was 1.2 × 107
mol/L (S/N = 3). The modified electrode exhibited good stability and can be easily regenerated. The relative standard deviation of the peak
current obtained for a 5.0 × 105 mol/L CAF solution was 3.0%. The influence of some important biological compounds, namely, ascorbic
acid, dopamine, and uric acid and addictive compounds like codeine, morphine, and acetaminophen on the CAF anodic peak current was
examined. The method was successfully applied for the determination of CAF in some practical samples.

Key words: caffeine; electrocatalysis; electrochemical determination; single-walled carbon nanotube; carbon-ceramic electrode

Caffeine (C8H10N4O2) is the common name for electrochemical determination of CAF [1,1828], but the
1,3,7-trimethylxanthine (3,7-dihydro-1,3,7-trimethyl-1H- electroanalysis of CAF has seldom been reported because the
purine-2,6-dione) [1,2]. Caffeine (CAF) is one of the most oxidation of CAF occurs at a very high positive potential.
widely used drugs in the world and has many important Carbon nanotubes are molecular scale tubes of graphitic
pharmacological effects such as the stimulation of central carbon with outstanding properties. They have attracted con-
nervous system, diuresis, and a positive effect on the cardio- siderable attention for their extraordinary structural, me-
vascular system [3,4]. CAF is also used therapeutically in chanical, electrical, and electrochemical properties [29]. Their
combination with ergotamine in the treatment of migraine and unique properties have led to applications in many fields such
in combination with nonsteroidal anti-inflammatory drugs in as electronics, medicine, aerospace industry, etc., which has
analgesic formulations [5]. It is sometimes included in anal- also prompted the need for analytical methods to characterize
gesic preparations because of its diuretic action [6]. Different and control the quality of CNTs. The use of CNTs as analytical
methods have been developed for the determination of CAF tools and the construction of nanodevices and nanosensors
[716]. Some of these methods, e.g., the chromatographic based on these materials are exciting areas of modern analyti-
methods, are time consuming, expensive, and need compli- cal science [3032].
cated pre-concentration or multiple solvent extraction and In the current work, an enhanced sensor based on sin-
trained technicians, but electrochemical methods are charac- gle-walled carbon nanotubes casted on a carbon-ceramic elec-
terized by simplicity, high sensitivity, good stability, low cost trode (SWCNT/CCE) was developed for the sensitive deter-
instrumentation, and onsite monitoring [17]. In order to en- mination of CAF. The electrochemical behavior of CAF
hance the sensitivity and stability of the measurement, various showed that the SWCNT/CCE exhibited obvious electrocata-
modified electrodes have been developed and used in the lytic activity for the oxidation of CAF because it greatly en-

Received 22 June 2012. Accepted 26 July 2012.

*Corresponding author. Tel: +98-412-4327541; Fax: +98-412-4327541; E-mail: [email protected]
Copyright © 2012, Dalian Institute of Chemical Physics, Chinese Academy of Sciences. Published by Elsevier BV. All rights reserved.
DOI: 10.1016/S1872-2067(11)60438-5
 Biuck HABIBI et al. / Chinese Journal of Catalysis, 2012, 33: 1783–1790

hanced the oxidation peak current of CAF with respect to that 1.4 Preparation of practical samples
from an unmodified CCE. The stability and electrocatalytic
activity and electroanalytical applications of the CAF-free drinks (mineral waters) were spiked with 0.5
SWCNT/CCE in electrooxidation and detection of CAF were mmol/L of CAF standard solution. The cola samples were
evaluated by different electrochemical techniques. Finally, in appropriately diluted (1:10) with the electrolyte to bring them
order to demonstrate the catalytic use of this modified elec- into the working range. In order to detect CAF in tea samples
trode for the electrooxidation of CAF in some practical sam- (Ahmad tea), a known amount (15.0 g) was weighted and
ples, we examined it with the voltammetric determination of dissolved in 250 ml of double distilled water followed by
CAF in tea, cola, and mineral water samples. boiling for 1 h on a hot plate with stirring. After allowing the
residue to settle, the hot solution was filtered and then used for
1 Experimental further experiments. A known amount (1 ml) of this solution
was added to 10 ml electrolyte solution of pH = 1.7. The
1.1 Chemicals quantitative determination of CAF was achieved by measuring
the oxidation peak current after background subtraction using
The SWCNT was manufactured by CNI (USA). Methyl- differential pulse voltammetry (DPV).
trimethoxysilane (MTMOS), caffeine, H2SO4, and other
chemicals were purchased from Merck or Fluka and used 2 Results and discussion
without further purification. Doubly distilled water was used in
the experiments. The pH of the solutions was adjusted with 0.1 2.1 Characterization of SWCNT/CCE
mol/L H2SO4 and NaOH.
Figure 1 is the SEM images of the SWCNT film on the CCE
1.2 Apparatus by SEM. Figure 1(a) shows the surface morphology of the bare
CCE immediately after polishing with emery paper of grade
The electrochemical experiments were carried out using a 1500. As seen in this image, the surface was dense, scaly, and
potentiostat/galvanostat (AUTOLAB PGSTAT-100) equipped had high porosity. Figure 1(b) shows the SEM image of the
with a USB interface and driven by a GPES 4.9 electro- same electrode after SWCNT casting on the surface of the
chemical software package (Eco Chemie, The Netherlands), a CCE. It can be seen that the SWCNT film was uniformly
three-electrode system and a personal computer for data stor-
age and processing. The three-electrode cell system employed
(a)
for the electrochemical studies composed of a saturated calo-
mel electrode (SCE) as the reference electrode, a platinum wire
as the auxiliary electrode, and the SWCNT modified CCE
(geometric surface area of 0.119 cm2) as the working electrode.
A copper wire was inserted through the composite end of the
working electrode to establish electrical contact. Scanning
electron microscopy (SEM) was performed on a LEO 440i
Oxford instrument.

1.3 Preparation of the bare and modified CCE

The bare CCE was prepared according to the procedure de-

(b)
scribed by Lev and coworkers [33,34] by mixing 0.15 ml
MTMOS, 0.30 ml methanol, and 10 μl hydrochloric acid (11
mol/L). This mixture was magnetically stirred for 2 min, after
which 0.3 g graphite powder was added and the resultant
mixture was shaken for an additional 1 min. A 5 mm length of
Teflon tube (34 mm inner diameter) was filled with the sol-gel
carbon mixture and dried under ambient conditions (25 ºC).
The dried CCE was first carefully polished with polishing
paper, rinsed with doubly distilled water, and finally dried by
air drying. To obtain a uniform layer of SWCNT on the CCE
surface, 20 μl of SWCNT (DMF/water = 1/3) solution (0.4 Fig. 1. SEM images of CCE surface immediately after polishing (a) and
mg/ml) was placed on the surface of the CCE and dried in air. after SWCNT casting on it (b).
 Biuck HABIBI et al. / Chinese Journal of Catalysis, 2012, 33: 1783–1790

coated on the electrode surface and formed a spaghetti-like 2.2 Electrochemical behavior of CAF on SWCNT/CCE
porous reticular formation. This surface morphology offered a
much larger surface area than the apparent geometric area. To Figure 3 shows the cyclic voltammetric responses of the bare
confirm this, the surface area of the modified electrode was CCE (curves 1 and 3) and the SWCNT/CCE (curves 2 and 4) in
estimated by cyclic voltammetry (CV) using Fe(CN)63/4 as the absence and presence of 3.0 × 103 mol/L CAF in 0.01
probe ions. The surface area was calculated using the mol/L, pH 1.7 H2SO4 at a scan rate of 50 mV/s. As previously
Randles-Sevcik equation [35]. The calculated electroactive reported [1,1820, 24,25,37,38], this oxidation system is
surface areas for the SWCNT/CCE were 0.330 ± 0.012 cm2, characterized by an anodic peak during the positive step and by
while the surface area for the bare CCE was 0.162 ± 0.018 cm2 the absence of any cathodic peak in the reverse scan, indicating
[36]. The significant increase in electroactive surface area that the oxidation is irreversible. As can be seen in Fig. 3(3),
suggested that the SWCNT/CCE is promising for electro- with the bare CCE, a small anodic current due to the weak
chemical sensing. In addition, the stability and reproducibility oxidation of CAF was observed, and no cathodic peak was
of the electrochemical behavior of the modified electrode were found, which indicated an irreversible heterogeneous charge
also investigated by the CV of Fe(CN)63/4 ions. The working transfer in this system. Under the same conditions, with the
stability of the modified electrode was verified by the elec- SWCNT/CCE a large anodic peak was observed (Fig. 3(4)).
trochemical behavior of Fe(CN)63/4 in successive sweeps of The anodic peak potential for the oxidation of CAF at the
CVs. The peak currents with 5.0 × 103 mol/L Fe(CN)63/4 modified CCE was 1.38 V versus the SCE, while that of CAF
(anodic peak at 0.205 V and cathodic peak at 0.121 V vs SCE) oxidation at the unmodified electrode was 1.48 V. A consid-
from the SWCNT/CCE surface was almost 98% of its initial erable decrease in the overpotential (100 mV) was observed
value after 200 cycles in the electrolyte solution and no sig- with the modified electrode. Note also that the background
nificant decrease was observed when the electrolyte used for current was slightly increased on the SWCNT/CCE due to the
200 repetitive cycles was replaced with a fresh solution (Fig. increase of the modified electrode surface area. The compari-
2). son of the CVs of CAF at the bare CCE and the SWCNT/CCE
Also, the storage stability of the chemically modified elec- (curves 3 and 4) shows that CAF can be effectively oxidized at
trode was very good. It was found to still have 98% of its initial the SWCNT/CCE with a higher peak current, which demon-
activity when kept in air at room temperature for more than 20 strated improved electron transfer kinetics and a promoted
d, and only 4% loss was found when the modified electrode oxidation current. The improvement of the electron transfer
was immersed in 0.1 mol/L sulfuric acid buffer solution for 48 kinetics and oxidation current from CAF at the SWCNT/CCE
h. In order to study the reproducibility of the electrode prepa- indicated that the SWCNTs had a good electrocatalytic effect
ration procedure, four independent CCEs were modified with on the electrooxidation of CAF. The high electrocatalytic ac-
SWCNTs. The cyclic voltammograms (CVs) of these modified tivity for the SWCNT-modified electrode was attributed to the
electrodes in 5.0 × 103 mol/L Fe(CN)63/4 solution were re- unique properties of the carbon nanotubes such as high specific
corded. The relative standard deviation (RSD) of the measured surface area, subtle electronic properties, and appropriate pore
cathodic peak currents was about 4% [36]. structure. In particular, the SWCNTs present an interlinked

390
Scan 1 Scan 100
100 (4)
Scan 200

50 290
(2)–(4) (1)
Current (A)

0
(3)
I/A

-50 190

-100
(1)
-150 90

-200 (2)

-0.2 0.0 0.2 0.4 0.6 0.8 1.0 -10

1.0 1.1 1.2 1.3 1.4 1.5 1.6
E/V vs SCE E/V vs SCE
Fig. 2. CVs of the SWCNT/CCE in the absence (a) and presence of 5.0 Fig. 3. CVs with the bare CCE (curves 1 and 3) and the SWCNT/CCE
× 103 mol/L Fe(CN)63/4 in 0.1 mol/L KCl at potential range 0.20.7 V (curves 2 and 4) in 0.01 mol/L, pH 1.7 H2SO4 in the absence and pres-
for the 1st (b), 100th (c), and 200th (d) cycles. Scan rate 50 mV/s, volume ence of 3.0 × 103 mol/L CAF. Scan rate 50 mV/s, volume of SWCNT
of SWCNT casted 20 µl. casted 20 µl.
 Biuck HABIBI et al. / Chinese Journal of Catalysis, 2012, 33: 1783–1790

highly mesoporous 3D structure with a larger electrochemi- stant, R is the rate gas constant, T is the absolute temperature,
cally accessible surface area and easier charge transfer at the К0 is the standard heterogeneous rate constant, and D is the
electrode/electrolyte interface. diffusion coefficient of CAF.
Further experiments were performed to study the transport The analysis of the data and plot of the corresponding curve
characteristics of CAF at SWCNT/CCE. The influence of the gave the peak potential Epa = 1.2734 + 0.0346lnυ (Epa in V, υ in
scan rate (υ) on the peak current of 3.0 × 103 mol/L CAF in V/s, R2 = 0.998), which corresponded to nα(1α) = 0.371. Be-
0.01 mol/L, pH 1.7 H2SO4 solution at the SWCNT/CCE was cause the electron transfer coefficient, α, for most electrode
investigated (Fig. 4). In the range 10 to 550 mV/s, the peak processes ranges from 0.7 to 0.3 [35], the electron transfer
current increased and the peak potential shifted to be more number (nα) should be one [41] and hence α was estimated to
positive. There was a good linear relationship between the peak be 0.63, which was in good agreement with an earlier result
current and the square root of the scan rate υ (Fig. 4 inset (a)). [41]. On the other hand, to confirm these results, another
The regression equation was Ipa = 0.51 + 80.33υ1/2 (R2 = 0.997), method was used to get the same information on the rate de-
suggesting that the electrode process was diffusion controlled. termining step. For this purpose, a Tafel plot was used (Fig. 4
This result was in agreement with data previously reported inset (b)) for the CAF at the surface of SWCNT/CCE using the
[1,18,20,22]. data derived from the rising part of the current-voltage curve.
We also used the above results to get information on the rate The slope of the Tafel plot is equal to nα(1α)F/2.3RT. Since
determining step. As can be seen in Fig. 4, when the scan rate this was 6.3622 V/decade, we obtained nα(1α) as 0.376. This
was increased, the anodic peak potential of CAF electrooxida- value again indicated that the one electron transfer process was
tion at the SWCNT/CCE was positively shifted slightly. For an the rate determining step and the transfer coefficient α = 0.624.
irreversible system, at high scan rates, Epa and lnυ should obey There was good agreement between the results from the two
the equation [39,40]: methods.
Epa = E' + RT/(1α)nαF × (0.780 + ln(DCAF1/2/К0) +
ln((1α)nαFυ/RT)1/2) 2.3 Influence of SWCNT thickness, electrolyte solutions
where α is the transfer coefficient, nα is the number of electrons and pH
transferred, υ is the potential scan rate, F is the Faraday's con-
The thickness of the SWCNT film depends on the amount of
SWCNT placed onto the electrode surface, and it influences the
1800
(a) 550 current. When the amount of SWCNT was increased, the peak
1600 1200 500 current increased, which resulted from the electrode surface
450 area increase. The peak current reached a maximum at 20 μl
I/A

600 and then remained almost unchanged up to 25 μl. When the

400
value exceeded 25 μl, the peak current decreased. This was due
0
1200 0 5 10 15 20 25 to the variation of film thickness and solidity. When the film
300
1/2/(mV/s)1/2 was thin, the amount of SWCNTs on the CCE was small and
250 the modification of electrode was not complete, and thus the
I/A

-3.8 (b) 200 peak current of CAF oxidation was small. When it was too
800 -4.2 150 thick, the conductivity of the film dropped, and the film no
LogI

longer adhered tightly, such that some SWCNTs were removed

-4.6 90 from the electrode surface. In addition, when the amount of
80
-5.0 70 SWCNT increased, the background current also increased,
1.1 1.2 1.3 1.4 60
400 50 thus, the signal/noise (S/N) ratio changed. Taking into account
E/V
40 the stability of the film and the S/N ratio, a 20 μL SWCNT
30
20 solution suspension was used for the following experiments.
Different electrolyte solutions can also influence the elec-
0 trochemical behavior of CAF. In order to investigate the effect
of the electrolyte [38,39], CVs of the SWCNT/CCE in the
1.0 1.2 1.4 1.6 presence of 3.0 × 103 mol/L CAF in 0.01 mol/L of different
E/V vs SCE
electrolytes including H2SO4, HNO3, and HCl were used. The
Fig. 4. CVs for electrooxidation of 3.0 × 103 mol/L CAF in 0.01mol/L highest current response was obtained with 0.01 mol/L H2SO4
H2SO4, pH 1.7 solution at the SWCNT/CCE at various scan rates. Plots of [39].
peak currents vs square root of scan rate (υ1/2) (inset (a)). Tafel plot for The pH value of the electrolyte is an important factor that
electrooxidation of CAF under the same condition with a scan rate of 10
affects the redox behavior of biomolecules and drugs. The peak
mV/s (inset (b)). Volume of SWCNT casted 20 µl.
potential and peak current of CAF are closely related to the pH
 Biuck HABIBI et al. / Chinese Journal of Catalysis, 2012, 33: 1783–1790

400 1.7 600

6
1.6
360 5

Icat/IL
1.5 450 4
320 3

E/V vs SCE
1.4
Ipa/A

2
300

I/A
280 1.3 1
1 2 3 4 5 6
1.2
t1/2/s1/2
240 (5)
1.1 150 (4)
(3)
200 1.0 (2)
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 (1)
pH 0
0 10 20 30 40 50 60
Fig. 5. Effect of pH on the peak potential and peak current for the elec- t/s
trooxidation of 3.0 × 103 mol/L CAF the SWCNT/CCE in H2SO4 solu-
tion. Volume of SWCNT was 20 µL. Fig. 6. Chronoamperograms of SWCNT/CCE in 0.01 mol/L H2SO4, pH
1.7 solution containing (1) 0 mol/L, (2) 5 × 105 mol/L, (3) 10 × 105
of the electrolyte solution. For the determination of the effect of mol/L, (4) 15 × 105 mol/L, and (5) 20 × 105 mol/L CAF. Plot of Icat/IL
versus t1/2 for 10 × 105 mol/L CAF (inset). Volume of SWCNT was 20
pH value on the SWCNT/CCE response, different pH values
µl.
were investigated by CVs. The effect of pH on the Epa and peak
current (Ipa) of CAF electrooxidation on the SWCNT/CCE was
examined (Fig. 5). different concentration of CAF. The transient currents decayed
In the pH range of 1.0 to 3.5, the Epa shifted positively with with time in a Cotrellian manner. The rate constant was ob-
increasing pH, and followed the linear equation Epa(V) = 1.19 + tained using the equation [43]:
0.058pH (R2 = 0.993). The value of dEp/dpH of 58 mV/pH for Icat/IL = π1/2(kcatc0t)1/2
CAF indicated the involvement of an equal number of protons where Icat and IL were the currents in the presence and absence
and electrons (4e, 4H+) in the electrooxidation reaction of of CAF, respectively, kcat is the rate constant (L/(mol·s)), c0 is
CAF [5,39,42]. On the other hand, Fig. 5 shows that the oxi- the bulk concentration (mol/L) of CAF, and t is the elapsed
dation peak current was dependent on pH. The current signal time (s). From the slope of Icat/IL versus t1/2 (inset of Fig. 6 for
decreased with increasing pH value [39]. It is evident that 1.0 × 105 mol/L of CAF), kcat for the electrooxidation of CAF
acidic media (pH < 2) were the most suitable. Therefore, 0.01 was calculated to be (1.8 ± 0.05) ×104 L/(mol·s).
mol/L H2SO4 solution and pH 1.7 were chosen for the subse-
quent experiments. 2.5 Analytical approaches
The mechanism for CAF electrooxidation (Scheme 1)
[21,3739,42] involves four electrons (4e) and four protons The analysis of the CAF solutions of different concentra-
(4H+). The first step is a 2e, 2H+ oxidation of the C-8 to N-9 tions was carried out under selected analytical conditions (pH =
bond to give the substituted uric acid. This is followed by an 1.7, amount of SWCNT suspension was 20 μl). There was a
immediate 2e, 2H+ oxidation to the 4,5-diol analog of uric good linear relationship in the cyclic voltammetric method
acid, which rapidly fragments. between the anodic peak current and the concentration for
concentrations from 5.0 × 106 mol/L to 5.0 × 103 mol/L (Fig.
2.4 Evaluation of the catalytic reaction rate constant 7). When the CAF concentration exceeded 5.0 × 103 mol/L,
the curve began to deviate from the linear plot. The detection
Chronoamperometry can be used to evaluate the catalytic limit was estimated to be 3.1 × 106 mol/L (S/N = 3).
reaction rate constant (kcat) for the electrochemical oxidation Differential pulse voltammetry (DPV) is often used to make
reaction between CAF and the redox sites of the sur- electrochemical measurements because of its high sensitivity
face-confined SWCNTs [4345]. Figure 6 shows chronoam- and low background current. In Fig. 8(a), there was a good
perograms of SWCNT/CCE in 0.01 mol/L H2SO4 solution in linear relationship between Ipa and CAF concentration in the

O O O
CH3 CH3 OHCH3
N + H2O N + 2H2O N
-
-

5
H3C N1 6
7 H3C N H3C N
2
8 O + 2H+ + 2e O + 2H+ + 2e
3 9
O N 4 N O N N O N N
OH
CH3 CH3 H CH3 H

Scheme 1. Electrooxidation mechanism of CAF.

 Biuck HABIBI et al. / Chinese Journal of Catalysis, 2012, 33: 1783–1790

500 Table 1 Comparison of the analytical performance of the present modi-

fied electrode with some other modified electrode for CAF determination
(17)
400 Detection
Dynamic liner
Modified electrode Method limit Ref.
range (μmol/L)
300 (μmol/L)
Nafion/MWNTs DPV 0.6400 0.23 1
I/A

BDDE DPV 0.583.0 0.035 5

200
Nafion/GCE DPV 0.99510.6 0.798 20
BDDE/Nafion® DPV 0.2120 0.1 23
100 MIP-CPE DPV 0.0625 0.015 23
(1)
MWNTs-Nafion/GCE DPV 2.945377.0 0.513 24
0 BDDE DPV 9.7110 7.0 27
1.0 1.1 1.2 1.3 1.4 1.5 1.6 PST/Nafion/GCE DPV 0.3100 0.1 28
E/V vs SCE Nafion-Gr/GCE DPV 0.440 0.12 36
DNA-CNTs/CPE DPV 0.51261.7 0.35 46
Fig. 7. CVs of CAF at concentrations of 0 (1), 20 × 106 (2), 50 × 106
6 6 6 6 Pt/CNTs/GCE DPV 10100 0.2 47
(3), 100 × 10 (4), 150 × 10 (5), 200 × 10 (6), 300 × 10 (7), 400 ×
6 6 6 6 6 SWCNT/CCE DPV 0.25100 0.12 This work
10 (8), 600 × 10 (9), 800 × 10 (10), 900 × 10 (11), 1100 × 10
(12), 1300 × 106 (13), 1700 × 106 (14), 1900 × 106 (15), 2000 × 106
(16), 2200 × 106 (17) mol/L in 0.01 mol/L H2SO4, pH 1.7 solution at the
SWCNT/CCS. Scan rate 50 mV/s, volume of SWCNT casted 20 µl. showed that the electrode gave good reproducibility. After
several days of repeated use, the response of the electrode was
decreased only slightly. This indicated that the electrode was
range 2.5 × 1071.0 × 107 mol/L (inset Fig. 8) with the re- very stable. A comparison of the analytical performance of the
gression equation: Ipa = 3.71 + 0.83cCAF (R2 = 0.996). The present modified electrode with some other relevant modified
detection limit was 1.2 × 107 mol/L (S/N = 3). electrodes for CAF determination is shown in Table 1. Com-
In ten successive determinations of a 5.0 × 105 mol/L CAF pared with the other electrochemical methods, the present
solution, the relative standard deviation (RSD) for the peak method has many advantages for the determination of CAF.
current was 3.0%. When a 5.0 × 105 mol/L CAF solution was Furthermore, the preparation of the electrode was simple, the
measured with the same electrode every several hours during modified electrode was quite stable, and it had good sensitivity
one day, the RSD of the peak current was 3.2% (n = 6), which and gave good reproducibility.

70
100
80 (d)
I/A

60 60
40
(a) 20
50
(17) 0
0 20 40 60 80 100
40 cCAF/mol/L) (5)
20 50
(5)
I/A

(c)
30 (b) 40
15 (4)
30
20 10 (3)
(1)
20
(1)
10 5 (2)
10
(1)
0 0 0
1.0 1.2 1.4 1.6 1.8 2.0 1.0 1.2 1.4 1.6 1.0 1.2 1.4 1.6
E/V vs SCE E/V vs SCE E/V vs SCE

Fig. 8. DPVs of CAF at concentrations of 0.25 (1), 0.32 (2), 0.42 3), 0.57 (4), 0.87 (5), 1.67 (6), 1.97 (7), 2.47 (8), 3.37 (9), 5.02 (10), 7.17 (11), 12.17
(12), 18.17 (13), 22.17 (14), 26.67 (15), 41.47 (16) and 60.17 × 106 (17) mol/L in 0.01 mol/L H2SO4, pH 1.7 solution (a). DPVs of cola (b) and tea (c)
samples results obtained after the addition of various amounts (0 (1), 10 (2), 20 (3), 35 (4), and 50 (5) μmol/L for cola and 0 (1), 10 (2), 40 (3), 100 (4), and
200 (5) μmol/L for tea) of standard CAF to samples (Pulse amplitude: 70 mV, pulse width: 50 ms, volume of SWCNT casted 20 µl). Plot of peak currents
vs CAF concentration (d).
 Biuck HABIBI et al. / Chinese Journal of Catalysis, 2012, 33: 1783–1790

Table 2 Determination of CAF in mineral water (CAF-free samples)

fabrication of a single-walled carbon nanotube modified car-
bon-ceramic electrode (SWCNT/CCE). The use of the re-
Spiked Found Recovery RSDb
Samplea versible couple [Fe(CN)6]3/4 as a probe electrochemical re-
(µmol/L) (µmol/L) (%) (%)
Zam Zam mineral water 1 50.00 49.25 98.5 3.7
action showed that the SWCNT/CCE has good chemical and
Zam Zam mineral water 1 50.00 49.60 99.2 3.5 electrochemical stability when it was used for the electrocata-
Vata mineral water 1 50.00 49.55 99.1 3.5 lytic oxidation of CAF. CAF was oxidized at the SWCNT/CCE
Vata mineral water 2 50.00 50.05 100.1 3.6 and produced an anodic peak at 1.38 V (vs SCE) in 0.01 mol/L,
a
Some available commercial mineral water brands in Iran. pH 1.7 H2SO4. The peak was suitable for the determination of
b
Number of Sample assayed = 5. the concentration of CAF. The peak currents in two different
voltammetric methods were linear in concentration over a large
concentration range. CAF concentration can be measured
2.6 Interference study without any interference in the presence of a large excess of
some important biological and addictive compounds such as
Possible interference of the detection of CAF with the
codeine, morphine, and acetaminophen. This procedure was
SWCNT modified CCE was investigated by the addition of
applied to the determination of CAF in some food samples.
various ions to the 0.01 mol/L, pH 1.7 H2SO4 solution with 5 ×
105 mol/L CAF. Common ions such as Na+, K+, Cu2+, Cl,
Acknowledgments
NO3, and CO32 did not show interference on CAF detection
even when these were present in over 500-fold excess con-
The authors gratefully acknowledge the research council of
centrations. To test interference in biological and other real
Azarbaijan University of Tarbiat Moallem for financial sup-
samples, 250-fold excess concentrations of ascorbic acid, do-
port.
pamine, uric acid, and vitamin B6, and 150-fold concentrations
of codeine, morphine, and acetaminophen were shown to give
References
no interference on the response current of CAF (signal change
< 5%), which showed that the present method has excellent 1 Yang S, Yang R, Li G, Qu L, Li J, Yu L. J Electroanal Chem,
selectivity towards CAF. 2010, 639: 77
2 Ashihara H, Suzuki T. Front Biosci, 2004, 9:1864
2.7 Application using practical samples 3 Rieg T, Steigele H, Schnermann J, Richter K, Osswald H, Vallon
V. J Pharmacol Exp Ther, 2005, 313: 403
CAF is the main and active ingredient in coffee, tea, and cola 4 Soloveychik V, Bin-Nun A, Ionchev A, Sriram S, Meadow W. J
nuts. Its concentration in CAF containing food samples is Perinatol, 2009, 29: 205
determined for quality control purposes. To simulate this, a 5 Lourenção B C, Medeiros R A, Rocha-Filho R C, Mazo L H,
known amount of the food sample solution was added to 10 ml Fatibello-Filho O. Talanta, 2009, 78: 748
6 Sawynok J. Drugs, 1995, 49: 37
of 0.01 mol/L and pH 1.7 H2SO4 solution, and their DPVs were
7 Baranska M, Proniewicz L M. Vib Spectrosc, 2008, 48: 153
recorded. First, the applicability of the present method was
8 Wei Y F, Wang D Y, Yan H T. Spectrosc Spect Anal, 2001, 21:
evaluated by analyzing CAF-free mineral water samples. The
534
results obtained are shown in Table 2. 9 Wang H, Chen L, Xu Y, Zeng Q, Zhang X, Zhao Q, Ding L.
It can be seen from the Table that the added and measured LWT-Food Sci Technol, 2011, 44: 1490
concentrations were generally in good agreement. The present 10 Soponar F, Mot A C, Sarbu C. Chromatographia, 2009, 69: 151
method was also applied to the analysis of cola and tea sam- 11 Wang A, Sun J, Feng H J, Gao S, He Z G. Chromatographia,
ples. The DPVs of cola (Zam Zam Cola) and tea (Ahmad tea) 2008, 67: 281
samples are presented in Figs. 8(b) and 8(c), respectively. The 12 Verenitch S S, Mazumder A. Anal Bioanal Chem, 2008, 391:
average value of four repeated measurements of the oxidation 2635
peak current was used to determine the concentration of CAF 13 Arinobu T, Hattori H, Kumazawa T, Lee X P, Mizutani Y, Katase
in the cola and tea samples to minimize possible effects due to T, Kojima S, Omori T, Kaneko R, Ishii A, Seno H. Forensic
Toxicol, 2009, 27: 1
the presence of other components in the samples. The amount
14 Aranda M, Morlock G. J Chromatogr Sci, 2007, 45: 251
of CAF in cola was found to be 35 ± 3.6 mg/L and in tea, it was
15 Emre D, Ozaltin N. J Chromatogr B, 2007, 847: 126
3.01 ± 0.19% (m/m), which were in good agreement with Ali-
16 Amini A, Barclay V, Rundlof T, Jonsson S, Karlsson A, Arvids-
zadeh’s results [23]. son T. Chromatographia, 2006, 63: 143
17 Sadik O A, Land W H, Wang J. Electroanalysis, 2003, 15: 1149
3 Conclusions 18 Zen J M, Ting Y S, Shih Y. Analyst, 1998, 123: 1145
19 Akyilmaz E, Turemis M. Electrochim Acta, 2010, 55: 5195
A simple, fast, and reproducible procedure was used for the 20 Brunetti B, Desimoni E, Casati P. Electroanalysis, 2007, 9: 385
 Biuck HABIBI et al. / Chinese Journal of Catalysis, 2012, 33: 1783–1790

21 Spãtaru N, Sarada B V, Tryk D A, Fujishima A. Electroanalysis, novich L, Sampath S. Chem Mater, 1997, 9: 2354
2002, 14: 721 35 Bard A J, Faulkner L R. Electrochemical Methods: Fundamentals
22 Martínez-Huitle C A, Suely Fernandes N, Ferro S, De Battisti A, and Applications. New York: Wiley, 2001
Quiroz M A. Diamond Relat Mater, 2010, 19: 1188 36 Habibi B, Jahanbakhshi M, Pournaghi-Azar M H. Microchim
23 Alizadeh T, Ganjali M R, Zare M, Norouzi P. Electrochim Acta, Acta, 2011, 172: 147
2010, 55: 1568 37 Dryhurst G, Hansen B H. J Electroanal Chem, 1971, 30: 407
24 Zhang J, Wang L P, Guo W, Peng X D, Li M, Yuan Z B. Int J 38 Telo J P, Vieira A J S C. J Chem Soc, Perkin Trans, 1997, 29:
Electrochem Sci, 2011, 6: 997 1755
25 Riahi S, Faridbod F, Ganjali M R. Sens Lett, 2009, 7: 42 39 Sun J Y, Huang K J, Wei S Y, Wu Z W, Ren F P. Colloids Surf B,
26 Katsu T, Tsunamoto Y, Hanioka N, Komagoe K, Masuda K, 2011, 84: 421
Narimatsu S. Anal Chim Acta, 2008, 620: 50 40 Zhang Z, Wang E. Electrochemical Principles and Methods. Bei-
27 Lourenção B C, Medeiros R A, Rocha-Filho R C, Fatibello-Filho jing: Science Press, 2000. 242
O. Electroanalysis, 2010, 22: 1717 41 Mersal G A M. Food Anal Methods, 2012, 5: 520
28 Guo S, Zhu Q, Yang B, Wang J, Ye B. Food Chem, 2011, 129: 1311 42 Goyal R N, Bishnoi S, Agrawal B. J Electroanal Chem, 2011,
29 Meyyappan M. Carbon Nanotubes Science and Applications. 655: 97
Florida: CRC Press LLC, 2005. 171 43 Galus Z. Fundamentals of Electrochemical Analysis. New York:
30 Valcárcel M, Simonet B M, Cárdenas S, Súarez B. Anal Bioanal Ellis Hornood, 1976. 10
Chem, 2005, 382: 1783 44 Ardakani M M, Karami P E, Rahimi P, Zare H R, Naeimi H.
31 Trojanowicz M. TrAC Trends Anal Chem, 2006, 25: 480 Electrochim Acta, 2007, 52: 6118
32 Merkoçi A. Microchim Acta, 2006, 152: 157 45 Laviron E. J Electroanal Chem, 1979, 101: 19
33 Tsionsky M, Gun G, Glezer V, Lev O. Anal Chem, 1994, 66: 46 Ly S Y, Lee C H, Jung Y S. Neuromol Med, 2009, 11: 20
1747 47 Wei Y, Zhang L, Shao C, Li C. Chem Anal (Warsaw), 2009, 54:
34 Lev O, Wu Z, Bharathi S, Glezer V, Modestov A, Gun J, Rabi- 607