1577

Full Paper

Electrochemistry and Adsorptive Stripping Voltammetric

Determination of Amoxicillin on a Multiwalled Carbon Nanotubes
Modified Glassy Carbon Electrode
Behzad Rezaei,* Sajjad Damiri
Department of Chemistry, Isfahan University of Technology, Isfahan, 84156-83111 I.R. Iran
*e-mail: [email protected]

Received: November 22, 2008

Accepted: March 15, 2009

Abstract
The study of electrochemical behavior of amoxicillin (AMX), a b-lactam antibiotic, is described on a multiwalled
carbon nanotubes (MWCNTs) modified electrode by electrochemical impedance spectroscopy (EIS) and adsorptive
stripping voltammetry for sensitive determination of AMX in pharmaceutical and human urine samples within a wide
pH range from 2.0 to 10.0. Also, studies by Fe2O3 nanoparticles modified carbon paste electrode show that iron oxide
impurities in the MWCNTs are not active sites for sensing of amoxicillin. Under optimized conditions, the oxidation
peak has two linear dynamic ranges of 0.6 – 8.0 and 10.0 – 80.0 mM with a detection limit of 0.2 mM and a precision of
< 4%.

Keywords: Multiwalled carbon nanotubes, Amoxicillin, Adsorptive stripping voltammetry, Fe2O3 nanoparticles,
Antibiotics, Nanotubes

DOI: 10.1002/elan.200804571

1. Introduction [11 – 23], and electrochemical methods [1 – 2, 24 – 28].

HPLC has been the most widely used technique for
Antibiotics of the b-lactam group are important antimicro- determination of this group of drugs, but presents some
bial agents that are widely used to treat infectious human disadvantages such as necessity to the large amount of high
and animal diseases and to enhance growth and yield in purity organic solvents, long equilibration and derivatiza-
agriculture. Amoxicillin (AMX), (D-a-amino-p-hydroxy- tion treatment. Moreover, the low solubility of amoxicillin
benzylpenicillin trihydrate, Scheme 1), is the only phenolic in the organic solvents becomes difficult to develop
penicillin and a moderate-spectrum b-lactam antibiotic used procedures based on extraction and pre-concentration
in humans and food-producing animals to treat several steps. Although, the electrochemical behavior of penicillins
diseases. b-lactam antibiotics presents a structure based on a has been extensively studied, papers about electrochemical
b-lactam ring responsible for the antibacterial activity and oxidation of amoxicillin describe very ill defined waves and
variable side chains that account for the major differences in high positive potential.
their chemical and pharmacological properties. AMX Carbon nanotubes (CNTs) continue to receive remark-
actives against a wide range of Gram-positive, and a limited able attention in electrochemistry [29 – 35]. Since their
range of Gram-negative organisms. It is usually the drug of discovery by Iijima [36] in 1991 using transmission electron
choice within the class because it is better absorbed, microscopy, CNTs have been the subject of numerous
following oral administration, than other b-lactam anti- investigations in chemical, physical and materials areas due
biotics. Amoxicillin is susceptible to degradation by b- to their novel structural, mechanical, electronic, and chem-
lactamase-producing bacteria, and so may be given with
clavulanic acid to decrease its susceptibility. In human urine,
amoxicillin is excreted in a ratio of 86  8%, its urine
concentration being 1.7  7 g L1 for the first 2 h after a
single oral dose of 500 mg [1 – 4].
The therapeutic importance of AMX required the devel-
opment of sensitive and rapid method for industrial quality
control and clinical monitoring. Various analytical methods
have been reported for the separation and determination of
amoxicillin based on spectroscopy [3 – 8], capillary electro-
phoresis [9, 10], high-performance liquid chromatography Scheme 1. Chemical structure of amoxicillin.

 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Electroanalysis 2009, 21, No. 14, 1577 – 1586
1578 B. Rezaei, S. Damiri

ical properties [37, 38]. Depending on their atomic structure, of nanoscale hematite affect their properties and potential
CNTs behave electrically as a metal or as a semiconductor applications [75, 76]. In the electrochemical sensors, up to
[39, 40]. The subtle electronic properties suggest that CNTs now, it has been reported that mesoporous Fe2O3 thin film
have the ability to promote charge-transfer reactions when electrodes catalyze glucose oxidation [77], tubelike a-Fe2O3
used as an electrode [41 – 44]. The modification of electrode shows unique sensing properties for nitric oxide [78], and
substrates with multiwalled carbon nanotubes (MWCNTs) mitochondrial cytochrome c (Mcc) sorbs strongly to hem-
for use in analytical sensing has been documented to result atite electrode from aqueous solution in a narrow pH range
in low detection limits, high sensitivities, reduction of due to opposite charge on Mcc and hematite [79].
overpotentials and resistance to surface fouling. MWCNTs The objective of the current work is to develop a sensitive
have been introduced as electrocatalysts [30 – 32, 45 – 46] and useful method for the determination of amoxicillin,
and CNTs modified electrodes have been reported to give based on the unusual properties of carbon nanotubes such as
super performance in the study of a number of biological strong adsorptive ability, huge specific area, subtle elec-
species, including cytochrome c [47], uric acid [48], NADH tronic properties as well as excellent electrocatalytic activ-
[49] and quercetin [50]. Recently, however, these reported ity. The electrochemical behavior of AMX on the multiwall
electrocatalytic properties of MWCNTs have been dispelled carbon nanotube-coated glassy carbon electrode (GCE)
[51, 52]. Compton and Banks group has demonstrated for was investigated by voltammetry and electrochemical
most electroactive species, via the comparison of MWCNT impedance spectroscopy techniques to better be determined
modified electrodes with edge plane pyrolytic graphite the interaction of AMX and the electrodes. The results
electrodes, that usually the electroactive sites of the showed that MWCNTs strongly enhanced the electron
MWCNTs are edge plane-like sites/defects which can occur transfer rate of amoxicillin oxidation and the determination
at the ends of the nanotubes or along the nanotube where sensitivity of amoxicillin was significantly improved. At the
tube compartments terminate [52, 53]. Accordingly MWCNTs coated GCE, the remarkable peak current
MWCNT modified electrodes are no more electrocatalytic enhancement and negative shift of peak potential occurred
than edge plane pyrolytic graphite electrodes except in a few for AMX oxidation, compared with that of a bare GCE.
cases where the observed response of the MWCNT Consequently, an ultrasensitive adsorptive stripping vol-
modified electrodes could not be mirrored with edge plane tammetric method based on the carbon nanotube-modified
pyrolytic graphite electrodes [54]. Specifically for the electrode was first developed for the determination of
electrochemical oxidation of hydrazine and the electro- amoxicillin in pharmaceutical and urine samples. This newly
chemical reduction of hydrogen peroxide, they have shown proposed method possesses following advantages such as
that the electrocatalysis is due to iron oxide impurities high sensitivity, rapid response, low cost and simplicity. Also,
rather than edge-plane-like sites/defects [54, 55] and the in the other section of this work, the activity of iron oxide
reduction of halothane is due to the presence of copper nanoparticles which are left in the nanotubes from produc-
oxide nanoparticles occluded within the carbon nanotubes tion process was studied on the a-Fe2O3 modified carbon
during their production [56]. These metal impurities have paste electrode to be recognized its capability in electro-
been found to never be completely removed using the catalytic oxidation of AMX.
common technique used to clean CNTs by stirring in strong
acid, even for excessively long exposure times and are
trapped in-between graphite sheets. In summary, the
2. Experimental
electroactivity of MWCNTs is a result of either edge-
plane-like sites/defects or metal impurities [45, 52, 54 – 56].
2.1. Apparatus and Reagents
On the other hand, above group has reported a new high-
purity metal-catalyst-free class of MWCNTs which are Electrochemical measurements were carried out in a con-
fabricated via a solid phase growth mechanism. This new ventional three-electrode cell, powered by an electrochem-
MWCNTs class will allow experimentalists to avoid any ical system comprising the Autolab system with PGSTAT 12
potential experimental misinterpretation, not only in elec- and FRA2 boards (Eco Chemie B.V., Utrecht, The Nether-
trochemistry and electrocatalysis, but also in related appli- lands). The system was run on a PC using GPES and FRA
cations [57]. 4.9 software. For impedance measurements, a frequency
Electrochemical properties of iron oxides have been range of 100 kHz to 10 mHz was employed. The AC voltage
widely studied [58 – 60] and they are of significant impor- amplitude used was 5 mV, and the equilibrium time was
tance in both technical and natural processes. a-Fe2O3 20 minutes. The MWCNTs modified GCE, a graphite
(hematite), which is the most stable iron oxide, with n-type electrode and a saturated Ag/AgCl reference electrode
semiconducting properties under ambient conditions, has was employed as a working, auxiliary and reference
been widely investigated in many technological fields for electrode, respectively. The MWCNTs were bought from
applications such as anode materials for lithium-ion cells Irans Research Institute of Petroleum Industry and synthe-
[61 – 64], gas sensors [65 – 69], catalysts [70 – 71], solar sized by chemical vapor deposition (CVD) with a diameter
energy conversion [72], electrochromism [73], and magnetic of 8 – 15 nm, a length 50 mM and the purity of 95%. The
materials [74, 75], owing to its low cost and high resistance to modified electrodes with carbon nanotube layers were
corrosion. It is well known that the particle sizes and shapes characterized by scanning electron microscopy (SEM) and

Electroanalysis 2009, 21, No. 14, 1577 – 1586 www.electroanalysis.wiley-vch.de  2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Determination of Amoxicillin 1579

energy dispersive X-ray (EDX) analysis. Alpha-Fe2O3 2.3. Preparation of Iron Oxide Nanoparticles Modified
nanoparticles (Particle size: 40 – 60 nm) purchased from Carbon Paste Electrode
Malek-Ashtar University of Technology. Spectrally pure
graphite powder (particle size < 50 mm) from Merck and The graphite powder and iron oxide nanoparticles mixture
high viscosity paraffin oil (density ¼ 0.88 g cm3) from Fluka was homogenized by ultrasonication in a beaker containing
were used for preparation of carbon paste electrode (CPE). 100 cc acetone. The modified carbon paste electrode was
Amoxicillin was purchased from Fluka. Capsules of amox- prepared by mixing 0.1 g of iron oxide nanoparticles with
icillin were prepared from Cosar Company (Tehran, Iran). 2.0 g of graphite powder and subsequently adding 0.250 g of
Each amoxicillin capsule contains amoxicillin trihydrate mineral oil. The final paste was obtained on evaporation of
equivalent to amoxicillin anhydrous 500 mg. Other non- the solvent. The modified carbon paste was packed into an
medicinal ingredients of capsule matrix are cellulose micro- electrode body, consisting of a plastic cylindrical tube (o.d.
crystalline, stearic acid and talc (magnesium silicate). 5 mm, i.d. 3 mm) equipped with a copper wire serving as an
0.002 M solutions of amoxicillin were freshly prepared in external electric contact. Appropriate packing was achieved
methanol-water (1 : 1, v : v) solution and were used as the by pressing the electrode surface against a filter paper.
stock solution.
Universal buffer (boric acid, phosphoric acid, acetic acid
and sodium hydroxide, 0.1 M) solutions with different pH
2.4. Preparation of Real Samples
values were used for the study of the pH influence. All
aqueous solutions were prepared with deionized water of Content of ten amoxicillin capsules labeled with amount of
resistivity not less than 18.0 MW at 25 8C. Prior to the 500 mg per capsule, were completely homogenized. 20 mg
adsorptive stripping voltammetry experiments, solutions of powders was accurately weighted and dissolved with
were stirred by a magnet for 5 minutes. ultrasonication in 100 cm3 of methanol-water (2 : 5, v : v)
solution. The resultant solution 10 times was diluted and
2 cm3 of the solution in 10 cm3 of 0.1 M buffer was used for
determination. The urine samples used for measurements
2.2. Preparation of MWCNT Suspension and Modified
were centrifuged and diluted 100 times with water without
GCE
any further pretreatment. The standard addition method
To eliminate metal oxide catalysts within the nanotubes, was used for the determination of amoxicillin in real
multiwalled carbon nanotubes were refluxed in the 2.0 M samples.
HNO3 for 15 hours, and then washed with twice-distilled
water and dried at room temperature. The purified
MWCNTs were dispersed in acetonitrile (0.1 mg MWCNTs
3. Results and Discussion
per 10 cc) by using ultrasonic agitation to obtain a relative
stable suspension. The GCE was carefully polished with
3.1. Characterization of MWCNT Modified Electrode
0.05 mm alumina slurry on a polishing cloth, and then
washed ultrasonically in methanol and water, respectively. The purity and dispersing state of the MWCNTs and iron
The cleaned GCE was coated by casting 300 mL of the black oxide nanoparticles were examined by SEM and energy
suspension of MWCNTs and dried in an oven at 60 8C. The dispersive X-ray (EDX) analysis. SEM micrographs showed
microscopic areas of the MWCNT-modified GCE and the most of MWCNTs are well dispersed in acetonitrile and no
bare GCE were obtained by CV using 1 mM K3Fe(CN)6 as a longer entwine together. EDX spectrum (not shown) similar
probe at different scan rates. For a reversible process, the to other reports [54], indicated the presence of significant
Randles – Sevcik formula has been used: amount of iron and sulfur after 15 hours purification in hot
nitric acid. The composition of the unwashed MWCNTs was
Ipa ¼ 2.69  (A s mol1 V1/2)n3/2AC0DR1/2v1/2 (1) 99.26%, C; 0.62%, Fe; 0.09%, S and 0.03%, Cl. By using
energy dispersive X-ray (EDX) analysis, the purified nano-
where Ipa (A) refers to the anodic peak current, n is the tubes showed near to three times lower content for iron and
electron transfer number, A (cm2) is the surface area of the sulfur. Banks and Comptons group has also shown the
electrode, DR (cm2 s1) is diffusion coefficient, C0 (mol cm3) electroactive metal impurities on/within the carbon nano-
is the concentration of K3Fe(CN)6, v (V s1) is the scan rate. tubes are not further removed by super washing and they
For 1 mM K3Fe(CN)6 in the 0.1 M KCl electrolyte: n ¼ 1 and may cause the observed “electrocatalysis” associated with
DR ¼ 7.6  106 cm2 s1, then from the slope of the ipa  v1/2 carbon-nanotube-modified electrodes [54 – 56].
relation, the microscopic areas can be calculated. On the
bare GCE, the electrode surface area was found 0.0314 cm2
and for MWCNT-modified GCE the microscopic area was
3.2. Electrochemical Behavior of Amoxicillin on GCE
nearly 2.9 – 3.0 times greater.
and MWCNT Modified Electrode
The voltammetric responses of 20 mM AMX on the bare and
MWCNTs modified electrode and in pH 7.5 and 3.0 have

 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim www.electroanalysis.wiley-vch.de Electroanalysis 2009, 21, No. 14, 1577 – 1586
1580 B. Rezaei, S. Damiri

Fig. 1. A) Stripping voltammograms of AMX at the GCE electrode, and B) background-subtracted stripping voltammogram at the
MWCNTs modified GCE. Other conditions: 20 mM AMX; scan rate of 100 mV s1; pH 7.5; and accumulation time of 6 minutes. Inset
shows corresponding background-subtracted stripping voltammograms in pH 3.0.

been compared at Figure 1. As can be seen, the oxidation of (A) and MWCNTs (B) electrodes recorded at 0.6 V as DC-
amoxicillin on the bare glassy carbon electrode and in acidic offset for 0.1 mM AMX in pH 7.5. Because impedance and
conditions shows an ill defined wave attributed to oxidation admittance quantities depends on the microscopic areas of
of the phenolic substituent to relevant carbonyl group. In the the electrodes and the surface area of the bare GC and
basic buffer solution, AMX does not show any electro- modified electrodes were different, therefore the impe-
oxidation signal on the GCE. The mentioned results, have dance elements were normalized with respect to the
been performed by adsorptive stripping measurements to be electrode surface area during data analysis after the experi-
enhanced the output signals. Under the similar conditions, ments. The Nyquist diagram of MWCNTs electrode repre-
amoxicillin yields a well-defined oxidation peak on the sent two overlapped, slightly depressed capacitive semi-
MWCNTs modified electrode at near to 1.0 and 0.6 V for circles. The high-frequencies semicircle is related to the
pH 3.0 and 7.5, respectively, and the oxidation peak current electron-transfer process and so the combination of charge-
significantly increases as compared with that at the bare transfer resistance and double-layer capacitance. The low-
GCE. The increase in the peak current and the lowering of frequencies semicircle can be related to an adsorption
oxidation potential are clear evidences of the catalytic effect process, which is the adsorption of the reactions intermedi-
of the MWCNTs toward the oxidation of amoxicillin. The ate(s) on the electrode surface. Also, bode and admittance
generated product of AMX oxidation is retained on the plots on Figure 3, clearly show the presence of one charge
electrode surface, which it can be identified by its reduction transfer step at the high frequencies and one adsorption
at þ 0.25 V at pH 3.0 in the reverse scan. phenomenon at the low frequencies [80].
Investigation of amoxicillin oxidation on the iron oxide However, the GC electrode represents a linear tail with a
nanoparticles modified carbon paste electrode, according to slope of near-unity at low frequencies and the charge
Figure 2, shows this material does not have any electro- transfer resistance of electrochemical oxidation process on
catalytic effect on the oxidation of amoxicillin. Therefore this electrode is very high. The linear tail was attributed to
iron oxide nanoparticles which are left in the nanotubes the semi-infinite diffusion of the electroreactant species
from production process can not be electroactive sites for [81]. The electron-transfer resistance, Rct, for the bare GC
oxidation of AMX and edge plane-like sites/defects of and MWCNTs electrode, equals to 410.0 and 40.9 W,
nanotubes, along with adsorption phenomenon, are respon- respectively. Therefore, the charge transfer rate for oxida-
sible for amplifying of the amoxicillin oxidation signal. tion of amoxicillin on the MWCNTs electrode surface, due
to their electrocatalytic effects, is significantly higher than
that of the GC electrode. It is interesting that admittance
diagram of MWCNTs electrode, shows a large diameter
3.3. Electrochemical Impedance Spectroscopy Studies
semicircle in high frequencies and a small admittance
Electrochemical impedance spectroscopy was also em- semicircle, related to adsorption of electrochemical reaction
ployed to investigate the oxidation of amoxicillin on both intermediate(s), at low frequencies. Moreover, electroox-
GC and MWCNTs electrodes. Figure 3 shows the bode plots idation of amoxicillin on the GCE shows only a small
and the Nyquist plots of the impedance surface densities (W/ admittance semicircle, indicating the low charge transfer
cm2) and admittance surface densities (S/cm2), on the GC rate on the electrode surface.

Electroanalysis 2009, 21, No. 14, 1577 – 1586 www.electroanalysis.wiley-vch.de  2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Determination of Amoxicillin 1581

Fig. 2. Electrochemical behavior of amoxicillin on the a) bare carbon paste electrode (CPE) without AMX; b) as (a) with 20 mM AMX;
c) a-Fe2O3 nanoparticles modified CPE without AMX; d) as (c) in the presence of 20 mM AMX. Other conditions scan rate of 100 mV
s1: pH 7.5 and accumulation time of 6 min. at OCP.

ZCPE ¼ (Y0jw)n (2)

The equivalent circuits compatible with the Nyquist
diagrams recorded for the GC and MWCNTs electrodes 1/2
ZW ¼ Y 1
0 (jw) (3)
are depicted in Schemes 2A and B, respectively. The model
equivalent circuits contain the solution resistance (Rs), the 1/2
ZT ¼ Y 1
0 (jw) coth[B (jw)1/2] (4)
charge-transfer resistance of amoxicillin oxidation (Rct) and
a constant phase element corresponding to the double-layer
capacitance (CPEdl). In addition, CPEads and Cads signify a where Y0 (the admittance parameter, S cm2 sn) and n
constant phase element and a capacitance representing the (dimensionless exponent) are two parameters independent
adsorption of the reaction intermediate(s) on the MWCNTs of frequency; j ¼ ( 1)1/2, w ¼ angular frequency ¼ 2pf, and
and GC electrodes, respectively. W, which is included in the B is the diffusion factor. ZCPE corresponds to the constant
circuit, is the Warburg element related to the semi-infinite phase angle element (CPE) impedance. Y0 ¼ Cdl only when
diffusion process on the GC electrode and T (hyperbolic n ¼ 1, and n is related to a (phase angle) by a ¼ (1  n) 90.
tangent element or bounded Warburg) shows the finite So, n ¼ 1 and a ¼ 0 stand for a perfect capacitor, and lower n
Warburg diffusion impedance of amoxicillin in the thin film values directly reflect the roughness of the electrode surface.
of MWCNTs. Impedance of CPE, W and T elements can be When n ¼ 0.5, it is equal to a Warburg impedance. When n ¼
expressed as [82]: 0, CPE is reduced to a resistor.
In the mentioned circuits, the charge-transfer resistance
of the electrode reaction is the only circuit element that has a
simple physical meaning describing how fast the rate of
charge transfer during electrooxidation of amoxicillin
changes with the electrode potential. The most widely
accepted explanation for the presence of constant phase
elements and the appearance of depressed semicircles in the
Nyquist plots is the microscopic roughness, causing an
inhomogeneous distribution in the solution resistance as
well as in the double-layer capacitance [83].
In this study, we used the above process on the MWCNTs
modified electrode for determination of amoxicillin in
pharmaceutical and urine samples and optimized the
various parameters that can affect on the amount of that
electrooxidation signal.
Scheme 2. The equivalent circuits compatible with the Nyquist
diagrams represented in Fig. 3.

 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim www.electroanalysis.wiley-vch.de Electroanalysis 2009, 21, No. 14, 1577 – 1586
1582 B. Rezaei, S. Damiri

Fig. 3. The Nyquist plots of the impedance surface densities (Z’’/cm2 vs. Z’/cm2) acquired for 0.1 mM AMX in pH 7.5 on the GC
electrode (A) and MWCNTs electrode (B). Bias in both diagrams was 0.6 V with 5 mV ac voltage amplitude and frequency range of
0.01 Hz to 100 kHz. Insets show their related a) bode plots and b) the admittance surface densities (Y’’/cm2 vs. Y’/cm2) plots. Points show
the experimental data and the full line is calculated from the optimized parameters.

3.4. Effect of Solution pH on the Peak Potentials and amoxicillin is protonated to give a cationic species. When
Peak Currents the pH is increased, in a first stage, the deprotonation of
carboxylic group (pKa1 ¼ 2.41) take places, yielding the
The peak potential and the peak current closely depend on zwitter ion species, existent in the pH range 4 – 6. In the next
the pH of buffer solution. Experimental results of 20 mM stages, the deprotonation of the amino group NH þ3 (pKa2 ¼
AMX in 0.1 M buffer solution at different pHs from 2.0 to 7.19) and phenolate group (pKa3 ¼ 9.38) are performed [84].
10.0 are shown in Figure 4. Direct oxidation of AMX By treatment of MWCNTs in nitric acid, carboxylic group
presents voltammetric signal on all pH range. When the pH can be created on the surface of nanotubes. Therefore in the
is increased it is observed that the peak potential shifts to acidic solutions, protonated AMX can not properly inter-
less positive values and shows some breaks in the slope at acted with working electrode surface and its oxidation
pHs 2.5 and 8.0 – 9.0. AMX presents three pKa values at current amount is relatively low. In the conditions that AMX
2.41; 7.19 and 9.38. In acidic medium, amino group of is deprotonated, there is better interaction with nanotubes

Electroanalysis 2009, 21, No. 14, 1577 – 1586 www.electroanalysis.wiley-vch.de  2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Determination of Amoxicillin 1583

are the units of the arguments. The relationship between the

peak potential and the scan rate (40 – 120 mV s1) was
studied. Based on the slope of the fitted line for the relevant
figure (not shown), with the equation: E (V) ¼ 0.0272  ln [v
(V s1)] þ 0.6384, R2 ¼ 0.9982, RT/2(1  a)nF ¼ 0.0272 V,
the value of n(1  a) was calculated to be 0.525. Also, the
slopes of tofel plots (E vs. log Ip) in low scan rates (10 –
50 mV s1) confirm this value of n(1  a). For n ¼ 1, a ¼
0.525, which indicates that the activation free energy curve
is approximately symmetrical in such an irreversible oxida-
tion process. On the other hand, according to the Equation 6
for totally irreversible process [85], a value in the optimized
conditions was found 0.58.

j Ep  Ep/2 j¼ 1.85 RT/aF ¼ (47.7/a) mV (6)

Fig. 4. Effect of pH on the peak current and peak potential at
the modified electrode. Other conditions, scan rate of 100 mV s1: From the mentioned results, the overall reaction mechanism
20 mM AMX and accumulation time of 6 min at OCP. can be attributed to the contribution of one electron and one
proton for oxidation of the phenolic group of amoxicillin to
generate the related quinone.
film and electrooxidation current is higher. The peak
current reaches maximum at pH 7.5, which was selected as
the optimized pH value for determination of amoxicillin.
3.6. Effects of Accumulation Potential and Time
The calibration curve of E versus pH, in the pH range of
3.0 – 8.0, has a slope of  44.4 mV pH1. This slope The oxidation peak currents of 20 mM AMX were measured
approximately suggests that the number of electron transfer by cyclic voltammetry after 6 min of accumulation at
is equal with that of hydrogen ions taking part in the different potential from  0.5 to þ 0.3 V in pH 7.5. It was
electrode reaction. observed the adsorptive stripping AMX response increases
rapidly upon increasing the accumulation potential. We also
observed the accumulation of AMX under open circuit
conditions, with a higher signal than that obtained at the
3.5. Effect of Scan Rate on the Peak Currents and Peak
mentioned accumulation potentials. Further experiments
Potentials
thus employed at open circuit potential.
The effects of scan rate on the peak current and peak In addition, the accumulation time affects significantly on
potential at the modified GCE in a 0.1 M universal buffer the adsorptive stripping voltammograms. By increasing the
were investigated by cyclic voltammetry in the presence of time, the oxidation peak current amplifies greatly at the first,
20 mM AMX at pH 7.5. The anodic peak current linearly and then reaches a steady amount after 6 minutes due to
increases as the scan rate increases, and the following adsorption saturation.
equation was obtained: I(mA) ¼ 1133.90  v (V s1) þ 23.43,
R2 ¼ 0.9961, indicating an adsorption controlled oxidation
process occurring at the modified GCE. The scan rate of
3.7. Performance of the System for Amoxicillin
100 mV s1 was selected as the optimum for amoxicillin
Measurements
determination. The linear relationship of peak potential (E)
and logarithmic scan rate (v) of an irreversible process obeys The calibration plot of amoxicillin determination is linear in
the following equation [85]; two concentration region of 0.6 – 8.0 and 10.0 – 80.0 mM with
detection limit of 0.2 mM (see Fig. 5). Using adsorptive
  1=2   stripping cyclic voltammetry under the optimum conditions
RT D were selected as: pH 7.5 (universal buffer) with a scan rate of
EðVÞ ¼ E ’ðVÞ þ ðVÞ 0:780 þ ln =ðs1=2 Þ
ð1  aÞnF k0 100 mV s1, and the stripping time of 6 min at open circuit
   potential. Equations of linear least square calibration curves
ð1  aÞnFn 1=2 1=2 over this ranges are:
þ ln =ðs Þ ð5Þ
RT I(mA) ¼ 14.67  CAMX  0.82 (R2 ¼ 0.9976) and I(mA) ¼
2.75  CAMX þ 127.96 (R2 ¼ 0.9964). Relative standard devi-
Here, E8’ is the formal potential, a the transfer coefficient, n ation (RSD) of < 4% for 20 mM amoxicillin (for 8 analysis)
the number of electrons involved in an electrode reaction, F showed excellent reproducibility. No obvious changes in the
the Faraday constant, DR the diffusion coefficient, k0 the anodic peak current were found for the same sample
standard heterogeneous rate constant, R the gas constant, T concentration when the modified electrode was kept under
the absolute temperature (300 K), and in the parentheses ambient conditions with the room temperature of about

 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim www.electroanalysis.wiley-vch.de Electroanalysis 2009, 21, No. 14, 1577 – 1586
1584 B. Rezaei, S. Damiri

Fig. 5. Calibration curves for determination of amoxicillin at the optimum conditions. Inset show the some of the raw voltammograms.

30 8C for one month or greater, thus, the modified electrode determination of AMX. 700-fold of Naþ, Kþ, NO 3 , Ca2þ,
  
is stable. Mg2þ, SO 2 4þ
4 , F , urea, 300-fold of NH , Cl , Br , CO 3 ,
2

The detection limit, linear dynamic range and sensitivity glucose, sucrose, lactose and fructose have almost no
for amoxicillin determination at this method, according to influence on the current response of AMX. All these
Table 1, are significantly better than other electrochemical indicated the peak current of amoxicillin is not affected by
methods reported by Biryol [2], Cavalheiro [27], and Zanoni all conventional cations, anions, and organic substances, but
[28]. Also the analysis can be performed on the wide pH the iron or manganese cations can be interfered.
range 2.0 – 10.0. As compared to most of the published
chromatographic and electrophoresis methods, which re-
quire lengthy and tedious extraction procedures, the pro-
3.9. Application
posed method does not require any filtration for analysis of
drugs from undissolved excipients and is rapid, very To evaluate the applicability of proposed method, the
sensitive and simple. Although the chromatographic proce- recovery of AMX was determined in the capsule and urine
dures has the advantage of simultaneous quantitation of the samples by adding the standard value of AMX to them. The
drug and its related substances when the chromatographic standard addition method was used for the analysis of
conditions are properly selected. prepared samples. The data given in Table 2 show the
satisfactory results for analytical determination of amox-
icillin in real samples.
3.8. Interference Studies
Under optimized experimental conditions described above, 4. Conclusions
the effects of some foreign species on the determination of
AMX at 20 mM level were evaluated in detail. The tolerance The results discussed above demonstrate that the electro-
limit was defined as the maximum concentration of the chemical response of amoxicillin by adsorption stripping
interfering substance that caused an error less than 3% for voltammetry on MWCNTs film can remarkably be en-

Table 1. Analytical parameters of several modified electrodes for AMX determination. LOD: limit of detection; LDR: linear dynamic
range; CPE: carbon paste electrode; SWV: square wave voltammetry.
Electrode Method LOD (mM ) LDR (mM ) Applicable pH range References
Modified CPE Cyclic voltammetry 0.812 10 – 200 < 2.3 [2]
Modified CPE SWV 8.49 18.9 – 91.9 5.5 [27]
Modified GCE SWV 0.92 2 – 25 < 7.0 [28]
MWCNTs modified GCE Cyclic voltammetry 0.2 0.6 – 8.0 and 10.0 – 80.0 2.0 – 10.0 This work

Electroanalysis 2009, 21, No. 14, 1577 – 1586 www.electroanalysis.wiley-vch.de  2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Determination of Amoxicillin 1585

Table 2. Determination of AMX in drug and urine samples in pH 7.5

No. Sample Added (mM ) Found [a] (mM ) Recovery (%) RSD (%)
Capsules
1 40 40.58  1.03 101.45 0.53
2 80 77.77  2.61 97.21 0.82
Urine
3 10 10.49  0.79 104.9 2.15
4 40 40.60  0.99 101.5 0.70

[a] Average of five replicate measurements

hanced in high pH range of 2.0 – 10.0 and iron oxide [14] K. M. Matar, Chromatographia 2006, 64, 255.
nanoparticles in the nanotubes which are left from produc- [15] N. Tavakoli, J. Varshosaz, F. Dorkoosh, M. R. Zargarzadeh, J.
tion process can not be active sites for amoxicillin oxidation. Pharm. Biomed. Anal. 2007, 43, 325.
[16] S. Riediker, R. H. Stadler, Anal. Chem. 2001, 73, 1614.
Furthermore, an adsorption process occurred during the
[17] W. Li, F. Tan, K. Zhao, J. Pharm. Biomed. Anal. 2006, 41,
redox process of AMX on the electrode surface; revealed by 594.
impedance measurements. Electrochemical studies for [18] D. M. Holstege, B. Puschner, G. Whitehead, F. D. Galey, J.
identifying of AMX oxidation mechanism confirm the Agric. Food Chem. 2002, 50, 406.
contribution of one electron and one proton in the process. [19] S. De Baere, M. Cherlet, K. Baert, P. De Backer, Anal.
Analysis by this method as compared to other reported Chem. 2002, 74, 1393.
electrochemical methods can be performed with a higher [20] K. H. Yoon, S. Y. Lee, W. Kim, J. S. Park, H. J. Kim, J.
Chromatogr. B, Anal. Technol. Biomed. Life Sci. 2004, 813,
sensitivity and determination range with low experimental 121.
detection limit of 0.2 mM. This modified electrode can be [21] S. De Baere, P. De Backer, Anal. Chim. Acta 2007, 586, 319.
properly used for determination of amoxicillin in drug and [22] S. M. Foroutan, A. Zarghi, A. Shafaati, A. Khoddamc, H.
human urine samples with satisfactory results. Movahed, J. Pharm. Biomed. Anal. 2007, 45, 531.
[23] A. Aghazadeh, G. Kazemifard, J. Pharm. Biomed. Anal.
2001, 25, 325.
[24] S. J. Lyle, S. S. Yassin, Anal. Chim. Acta 1993, 274, 225.
5. Acknowledgements [25] İ. Biryol, B. Uslu, Z. KuÅukyavuz, S.T.P. Pharma Sci. 1998, 8,
383.
The authors wish to thank Isfahan University of Technology [26] B. Uslu, İ. Biryol, J. Pharm. Biomed. Anal. 1999, 20, 591.
(IUT) Research Council and Center of Excellence in sensor [27] M. F. Bergamini, M. F. S. Teixeira, E. R. Dockal, N. Bocchi,
(IUT) for support of this work. E. T. G. Cavalheiro, J. Electrochem. Soc. 2006, 153, E94.
[28] D. P. Santos, M. F. Bergamini, M. V.B. Zanoni, Sens. Actua-
tors B 2008, 133, 398.
[29] Q. Zhao, Z. Gan, Q. Zhuang, Electroanalysis 2002, 14, 1609.
6. References [30] A. MerkoÅi, Microchim. Acta. 2006, 152, 157.
[31] A. MerkoÅi, M. Pumera, X. Llopis, B. Pérez, M. del Valle, S.
[1] L. Koprowski, E. Kirchmann, L. E. Welch, Electroanalysis Alegret, Trends Anal. Chem. 2005, 24, 826.
1993, 5, 473. [32] J. J. Gooding, Electrochim. Acta 2005, 50, 3049.
[2] B. Uslu, İ. Biryol, J. Pharm. Biomed. Anal. 1999, 20, 591. [33] A. MerkoÅi, Electroanalysis 2007, 19, 739.
[3] A. Garca-Reiriz, P. C. Damiani, A. C. Olivieri, Talanta 2007, [34] J. Wang, Electroanalysis 2005, 17, 7.
71, 806. [35] M. Trojanowicz, Trends Anal. Chem. 2006, 25, 480.
[4] A. Garca-Reiriz, P. C. Damiani, A. C. Olivieri, Anal. Chim. [36] S. Iijima, Nature 1991, 354, 56.
Acta 2007, 588, 192. [37] P. M. Ajayan, Chem. Rev. 1999, 99, 1787.
[5] G. G. Mohamed, J. Pharm. Biomed. Anal. 2001, 24, 561. [38] T. W. Odom, J. L. Huang, P. Kim, C. M. Lieber, Nature 1998,
[6] A. Pasamontes, M. P. Callao, Anal. Chim. Acta 2004, 515, 391, 62.
159. [39] J. W. Jang, D. K. Lee, C. E. Lee, T. J. Lee, C. J. Lee, S. J. Noh,
[7] H. Salem, Anal. Chim. Acta 2004, 515, 333. Solid State Commun. 2002, 122, 619.
[8] M. Q. Al-Abachi, H. Haddi, A. M. Al-Abachi, Anal. Chim. [40] D. Tekleab, R. Czerw, D. L. Carroll, P. M. Ajayan, Appl.
Acta 2005, 554, 184. Phys. Lett. 2000, 76, 3594.
[9] S. M. Santos, M. Henriques, A. C. Duarte, V. I. Esteves, [41] J. M. Nugent, K. S. V. Santhanam, A. Rubio, P. M. Ajayan,
Talanta 2007, 71, 731. Nano Lett. 2001, 1, 87.
[10] G. Pajchel, K. Pawłowski, S. Tyski, J. Pharm. Biomed. Anal. [42] K. Gong, Y. Yan, M. Zhang, L. Su, S. Xiong, L. Mao, Anal.
2002, 29, 75. Sci. 2005, 21, 1383.
[11] V.Gamba, G. Dusi, Anal. Chim. Acta 2003, 483, 69. [43] J. Koehne, J. Li, A. M. Cassell, H. Chen, Q. Ye, H. Tee Ng, J.
[12] J. I.D. Wibawa, D. Fowkes, P. N. Shaw, D. A. Barrett, J. Han, M. Meyyappan, J. Mater. Chem. 2004, 14, 676.
Chromatogr. B, Anal. Technol. Biomed. Life Sci. 2002, 774, [44] M. Musameh, N. S. Lawrence, J. Wang, Electrochem. Com-
141. mun. 2005, 7, 14.
[13] G. Hoizey, D. Lamiable, C. Frances, T. Trenque, M. [45] C. E. Banks, R. G. Compton, Analyst 2006, 131, 15.
Kaltenbach, J. Denis, H. Millart, J. Pharm. Biomed. Anal. [46] J. Wang, M. Musameh, Y. Lin, J. Am. Chem. Soc. 2003, 125,
2002, 30, 661. 2408.

 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim www.electroanalysis.wiley-vch.de Electroanalysis 2009, 21, No. 14, 1577 – 1586
1586 B. Rezaei, S. Damiri

[47] G. Zhao, Z. Yin, L. Zhang, X. Wei, Electrochem. Commun. [67] Y. Wang, J. Cao, S. Wang, X. Guo, J. Zhang, H. Xia, S. Zhang,
2005, 7, 256. S. Wu, J. Phys. Chem. C 2008, 112, 17804.
[48] S. Wei, F. Zhao, B. Zeng, Microchim. Acta 2005, 150, 219. [68] Y. Wang, Y. Wang, J. Cao, F. Kong, H. Xia, J. Zhang, B. Zhu,
[49] M. Musameh, J. Wang, A. MerkoÅi, Y. Lin, Electrochem. S. Wang, S. Wu, Sens. Actuators B 2008, 131, 183.
Commun. 2002, 4, 743.
[50] J. He, X. Lin, J. Pan, Electroanalysis 2005, 17, 1681. [69] Z. Sun, H. Yuan, Z. Liu, B. Han, X. Zhang, Adv. Mater. 2005,
[51] R. R. Moore, C. E. Banks, R. G. Compton, Anal. Chem. 17, 2993.
2004, 76, 2677. [70] A. Brown, J. Hargreaves, B. Rijniersce, Catal. Lett. 1998, 53, 7
[52] C. E. Banks, R. R. Moore, T. J. Davies, R. G. Compton, [71] P. Li, D. E. Miser, S. Rabiei, R. T. Yadav, M. R. Hajaligol,
Chem. Commun. 2004, 16, 1804. Appl. Catal. B 2003, 43, 151
[53] C. E. Banks, R. G. Compton, Analyst 2005, 130, 1232. [72] X. Qian, X. Zhang, Y. Bai, T. Li, X. Tang, E. Wang, S. Dong,
[54] C. E. Banks, A. Crossley, C. Salter, S. J. Wilkins, J. Nanopart. Res. 2000, 2, 191.
R. G.Compton, Angew. Chem. Int. Ed. 2006, 45, 2533. [73] G. Zotti, G. Schiavon, S. Zecchin, J. Electrochem. Soc. 1998,
[55] B. Šljukić, C. E. Banks, R. G. Compton, Nano Lett. 2006, 6, 145, 385.
1556. [74] R. D. Zysler, D. Fiorani, A. M. Testa, J. Magn. Magn. Mater.
[56] X. Dai, G. G. Wildgoose, R. G. Compton, Analyst 2006, 131, 2001, 224, 5
901. [75] C. Wu, P. Yin, X. Zhu, C. OuYang, Y. Xie, J. Phys. Chem. B
[57] C. P. Jones, K. Jurkschat, A. Crossley, R. G. Compton, B. L. 2006, 110, 17806.
Riehl, C. E. Banks, Langmuir 2004, 23, 9501. [76] S. Zeng, K. Tang, T. Li, J. Coll. Inter. Science 2007, 312, 513.
[58] X. Wang, L. Gao, H. Zheng, M. Ji, T. Shen, Z. Zhang, J. [77] J. Gong, L. Wang, K. Zhao, D. Song, L. Zhang, Electrochem.
Cryst. Growth 2004, 269, 489. Commun. 2008, 10, 1222.
[59] T. Grygar, F. Marken, U. Schrçder, F. Scholz, Collec. Czech. [78] C. M. Eggleston, N. Khare, D. M. Lovelace, J. Elect. Spec.
Chem. Commun. 2002, 67, 163. Relat. Phen. 2006, 150, 220.
[60] C. Y. Cummings, M. J. Bonné, K. J. Edler, M. Helton, A. [79] B. T. Hang, H. Hayashi, S. H. Yoon, S. Okadac, J. Yamaki, J.
McKee, F. Marken, Electrochem. Commun. 2008, 10, 1773. Power Sources 2008, 178, 393.
[61] J. Chen, L. Xu, W. Y. Li, X. L. Gou, Adv. Mater. 2005, 17, 582. [80] S. Majdi, A. Jabbari, H. Heli, H. Yadegari, A. A. Moosavi-
[62] X. L. Gou, G. X. Wang, J. S. Park, H. Liu, J. Yang, Nano- Movahedi, S. Haghgoo, J. Solid State Electrochem., in press.
technology 2008, 19, 125606. [81] E. Barsoukov, J. R. Macdonald, Impedance Spectroscopy:
[63] H. Liu, G. Wang, J. Park, J. Wang, H. Liu, C. Zhang, Theory, Experiment, and Applications, 2nd ed., Wiley, New
Electrochim. Acta 2009, 54, 1733 York 2005.
[64] M. V. Reddy, T. Yu, C. H. Sow, Z. X. Shen, C. T. Lim, G. V. S. [82] User Manual for Software FRA 4.9 (Autolab system), Eco
Rao, B. V. R. Chowdari, Adv. Funct. Mater. 2007, 17, 2792 Chemie B.V., The Netherlands 2007.
[65] L. H. Huo, Q. Li, H. Zhao, L. J. Yu, S. Gao, J. G. Zhao, Sens. [83] A. Maritan, F. Toigo, Electrochim. Acta 1990, 35, 141.
Actuators B 2005, 107, 915. [84] M. M. Shoukry, Talanta 1992, 39, 1625.
[66] O. K. Tan, W. Cao, W. Zhu, J. W. Chai, J. S. Pan, Sens. [85] A. J. Bard, L. R. Faulkner, Electrochemical Methods: Funda-
Actuators B 2003, 93, 396. mentals and Applications, 2nd ed., Wiley, New York 2001,
p. 236.

Electroanalysis 2009, 21, No. 14, 1577 – 1586 www.electroanalysis.wiley-vch.de  2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim