- Email: [email protected]
Very sensitive electrochemical sensor for moniliformin detection in maize samples
Accepted Manuscript Title: Very sensitive electrochemical sensor for moniliformin detection in maize samples Author: Paulo C´esar D´ıaz Toro Fernando Javier Ar´evalo Mar´ıa Ver´onica Fumero Mar´ıa Alicia Zon H´ector Fern´andez PII: DOI: Reference:
S0925-4005(15)30623-7 http://dx.doi.org/doi:10.1016/j.snb.2015.11.042 SNB 19300
To appear in:
Sensors and Actuators B
Received date: Revised date: Accepted date:
9-6-2015 15-10-2015 7-11-2015
Please cite this article as: P.C.D. Toro, F.J. Ar´evalo, M.V. Fumero, M.A. Zon, H. Fern´andez, Very sensitive electrochemical sensor for moniliformin detection in maize samples, Sensors and Actuators B: Chemical (2015), http://dx.doi.org/10.1016/j.snb.2015.11.042 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ip t
Very sensitive electrochemical sensor for moniliformin detection in maize samples
Paulo César Díaz Toroa, Fernando Javier Arévaloa, María Verónica Fumerob, María Alicia Zona*
Departamento de Química. Universidad Nacional de Río Cuarto. Agencia Postal N° 3.
an
a
us
cr
and Héctor Fernándeza*
(X5804ALH) - Río Cuarto. Fax: 54-358-4676233; Tel: 54-358-4676233; Argentina. Departamento de Microbiología e Inmunología. Universidad Nacional de Río Cuarto.
M
b
Agencia Postal N° 3. (X5804ALH) - Río Cuarto. Fax: 54-358-4676231; Tel: 54-358-
Ac ce pt e
d
4676231. Argentina.
* Correspondig Authors
E-mail addresses: [email protected] (P. C. Díaz Toro); [email protected] (F. J. Arévalo); [email protected] (M. V. Fumero); [email protected] (M. A. Zon); [email protected] (H. Fernández).
1
Page 1 of 34
Abstract A highly sensitive electrochemical sensor based on self assembled monolayers (SAMs) of cysteamine on gold electrodes (Au-CA) in pH 4 citrate buffer solutions was
ip t
developed to quantify the moniliformin (MON) mycotoxin by cyclic voltammetry in contaminated maize samples. Parameters of SAMs modified gold electrodes such as
cr
concentration of thiol, modification time and MON accumulation time were adjusted to
us
obtain the best performance of the sensor. The calibration curve was linear in the MON concentration range from 1 x 10-9 to 1 x 10-7 mol L-1. A limit of detection of 8.3 x 10-10 mol
an
L-1 (0.1 ppb) was obtained. The standard addition method was used to minimize matrix effects. The concentration of MON determined with the electrochemical sensor was in very
M
good agreement with results found by HPLC UV-Vis. The sensor showed a good analytical
Ac ce pt e
d
performance with a relative standard deviation (RDS%) of about 3%.
Keywords: Mycotoxins; Moniliformin; Self-assembled monolayers; Cyclic voltammetry; Maize samples.
2
Page 2 of 34
1. Introduction
Moniliformin (3-hydroxy-3-cyclobutene-1,2-dione, MON, Fig. 1) is a mycotoxin produced by different species of Fusarium, mainly F. proliferatum [1]. In the nature, MON
ip t
occurs as the sodium or potassium salt, and presents a pKa less than 1, which makes it soluble in polar solvents [1].
cr
The mechanism of toxicity of MON is not yet well known, and its toxic-kinetics is
us
still unknown. MON was found in cereals such as maize, rice, barley, wheat, oats, rye and triticale [2], although the greatest contamination was found in maize samples [3]. The
an
International Agency for Researcher on Cancer (IARC) classifies MON in the Group 3, because the agent is not classifiable as carcinogenic to humans [4]. Several studies have
M
confirmed the toxic effect produced by MON in animals [5,6] and its potential responsibility in the Keshan disease that occurred in China [7]. It has been proposed that
d
MON inhibits the oxidation of intermediates in the tricarboxylic acid cycle [8,9], causing
Ac ce pt e
respiratory diseases, and myocardial degeneration in animals, and leading, in some cases, death. It is also proposed that MON competes with pyruvate for active sites of α– ketoglutarate dehydrogenase, pyruvate decarboxylase, and aceto-hydroxy acid synthetase enzymes [10]. Thus, MON was classified into a group of emerging mycotoxins, highlighting its potential toxic effects on both animals and humans [11]. Until now, there is no regulations respect the maximum levels of MON allowed in food or grains, or even an official technique for its quantification [2, 11]. Thus, the development of relatively fast analytical methods to determine MON in contaminated samples is a major challenge, where sensitivity, selectivity, precision, and accuracy are required. However, several methods have been developed to quantify MON based mostly on chromatography techniques,
3
Page 3 of 34
especially the high performance liquid chromatography (HPLC). All these techniques require a pre-treatment of the samples to achieve a good performance. Often, derivatization reactions or ion-pair agents are necessary to obtain products that can be detected by HPLC
ip t
UV-Vis [12,13], HPLC fluorescence [14], TLC-UV [15], LC-MS/MS [16]. The technique of capillary zone electrophoresis with diode array detection (CZE-DAD) uses a similar pre-
cr
treatment to that of HPLC [17]. The limits of detection (LOD) of these techniques are
us
higher than of 40 ppb. Gilbert et al [18] have developed a methodology to detect MON by GC-MS and found a LOD of 0.5 ppb. The major problem of this method is the
an
derivatization step of MON with N-methyl-N-(tert-butyldimethylsilyl) trifluroacetamide containing 1 % tert-butyldimethylchlorosilane. We have recently determined the
M
thermodynamic and kinetics parameters of MON oxidation at glassy carbon electrodes in acetonitrile solution [19]. We have also proposed an electroanalytical methodology to
Ac ce pt e
satisfactory.
d
detect MON in this medium, but the determination of MON in maize samples was not
The use of self assembled monolayers (SAMs) to functionalize clean metallic
surfaces (either obtained by soaking the surfaces in solutions or in vapour phase) has been widely used in the development of electroanalytical techniques to quantify of a large number of analytes [20]. These SAMs are characterized by their high degree of orientation and stability due to the thiol is covalently bonded to the corresponding metallic substrate, particularly, gold, platinum, and mercury [21]. In addition, different interaction forces are present among adsorbed molecules, such as Van der Waals forces, hydrogen bond, and πinteractions [21]. Organosulfur compounds can be aliphatic or aromatic, and have different chain length and variable tail functional groups. Tail groups can be chosen to achieve a better affinity and specificity towards a given analyte, and provide selectivity and 4
Page 4 of 34
sensitivity to the sensor. However, the final response depends on the alkane-thiol chain length and/or the nature of tail functional groups [21,22]. Short alkane-thiol SAMs form thin monolayers with high defect sites generating pinholes [23], where the electrocatalytic
ip t
activity of different substrates has been observed [24]. The major advantages of SAMs with respect to other surface modification is the well defined control over the composition,
cr
structure, thickness and orientation of the monolayer, which is given by the nature and the
us
alkane-thiol chain length [25,26]. On the other hand, monolayers from organosulfur compounds spontaneously self-assembled onto clean metallic surfaces are easy and simple
an
to form and require not special equipments [21].
Because of its low pKa, MON is present in its deprotonated form in pH 4 buffer
M
solutions, while cysteamine (pKa ≅ 7.6) adsorbed on the gold electrode is mainly protonated at pH 4 [27]. Thus, in principle, the electrostatic interaction between MON and the gold
d
electrode modified with cysteamine SAMs can be used to pre-concentrate MON, and
Ac ce pt e
electro-catalyze its electrochemical oxidation.
In this work, we describe the development of a very sensitive electrochemical
sensor to determine MON in maize samples, which is based on MON oxidation on cysteamine SAMs modified polycrystalline gold electrodes (Au-CA) in pH 4 citrate buffer solutions (CBS). The electrochemical technique used was cyclic voltammetry (CV). The results obtained by the electrochemical method were compared with those obtained by HPLC, indicating a very good performance of the proposed electrochemical sensor. As far as we know, this is the first electrochemical sensor reported for the detection of MON in real samples.
5
Page 5 of 34
2. Materials and methods 2.1. Reagents All reagents were of analytical grade. Sodium salt of MON, 2-aminoethanethiol
ip t
(cysteamine, CA), 2-(diethylamino) ethanethiol hydrochloride (2-DAET), 8-amino-1octanethiol hydrochloride (8-AOT), 4-amino-thiophenol (4-ATP) and 4-mercaptopyridine
cr
(4-MP) were purchased from Sigma-Aldrich (St. Louis, USA).CBS, ethanol, potassium
us
dihydrogen phosphate (KH2PO4), tetrabutylammonium hydroxide (TBAOH), phosphoric acid, sulfuric acid and hydrogen peroxide were purchased from Merck p.a. (Darmstadt,
Argentina). All reagents were used as received.
an
Germany). Acetonitrile and methanol were HPLC grade (Sintorgan, Buenos Aires,
M
A MON stock solution (4.2 x 10-3 mol L-1) was prepared in H2O (Mili-Q) and stored at 4ºC. Working solutions were prepared daily by adding aliquots of the stock solution to
d
CBS. The final concentration of MON was controlled by UV-Vis spectroscopy (see below).
Ac ce pt e
Solutions of thiols were prepared daily in ethanol at different concentrations and stored in darkness.
The mobile phase used for HPLC experiments was an ion pair buffer (10 mL of ion
pair solution and 50 mL of acetonitrile were diluted with water to 1 L). The ion pair solution was prepared by mixing 50 mL of 20 % (w/w) TBAOH in water and 100 mL of 1.1 mol L-1 KH2PO4. The pH was adjusted to 7.0 by addition of phosphoric acid solution (Sigma-Aldrich, St. Louis, USA).
Both uncontaminated and contaminated with Fusarium temperatum RCFT914 strain [28] maize samples were provided by the Departamento de Microbiología e Inmunología, Facultad de Ciencias Exactas, Físico-Químicas y Naturales, Universidad Nacional de Río Cuarto. 6
Page 6 of 34
2.2. Materials and apparatus Cyclic voltammetry (CV) measurements were carried out in a conventional threeelectrode cell. The working electrode was a polycrystalline gold disk (CH Instrument,
ip t
Austin, USA) of 2 mm of diameter. Previous to perform the experiments, the gold electrode was successively polished on BAS cloth with wet alumina powder (0.3 and 0.05 μm from
cr
Fischer, respectively), washed with water and cleaned in a H2SO4 + 30% H2O2 (3:1 v/v)
us
solution during 3 min. Then, it was placed in an ultrasonic bath for 5 min. Finally, it was cycled in 0.5 mol L-1 H2SO4 between 0.2 and 1.7 V until a typical voltammogram of a
an
polycrystalline Au clean surface was obtained [29]. The electrochemical area of the gold electrode was calculated through the charge of the cathodic peak for the reduction of their
M
oxides assuming a charge density of 420 μC cm-2, which corresponds to the reduction of one monolayer of gold oxide for polycrystalline gold [30]. A roughness factor of 5.4 ± 0.7
d
was calculated from the ratio of electrochemical and geometric surface areas and it was
Ac ce pt e
kept constant in all experiments [27]. A similar value of roughness factor was also found by other authors, who used a similar extensive pre-treatment prior to the modification of electrodes [31,32]. An Ag/AgCl (BAS, West Lafayette, USA) electrode and a Pt wire of large area (A ≈ 2 cm2) were used as the reference and auxiliary electrodes, respectively. Cyclic voltammograms were performed with a Micro AutoLab PGSTAT 101 potentiostat (Eco-Chemie, Utrecht, The Netherlands) controlled by the NOVA 1.10.2 software. All measurements were carried out at 25.0 ± 0.2 ºC. UV–Vis absorption spectra were recorded immediately after the preparation of MON solutions to determine their concentrations. A spectrophotometer Hewlett–Packard model 8452A was used. The UV-Vis measurements were performed using silica cells of 1 cm path length. Absorption spectra of MON were recorded at different concentrations in 7
Page 7 of 34
CBS. MON absorption spectra showed two bands, with maxima at λ = 227 nm and at λ = 258 nm, respectively. Plots of A vs c*MON at λ = 227 and λ = 258 nm were linear from 8.3 x
103 mol-1 L cm-1 and ε258 = (8.02 ± 0.05) x 103 mol-1 L cm-1.
ip t
10-6 to 4.5 x 10-5 mol L-1. Molar extinction coefficients in CBS were ε227 = (25.46 ± 0.05) x
A SAX Strong Anion Exchange column was used for clean-up of maize samples
cr
before CV and HPLC measurements. The HPLC measurements were performed with a
us
Hewlett Packard HPLC Serie 1100, coupled to UV-Vis detector. The column used was a Hypersil C18 (150 x 4.2 mm, 5 μm; Phenomenex) and a guard column ODS-Hypersil (20 x
an
2.1 mm, 5 μm; Phenomenex). A flow rate of 1 mL min-1 and an injection volume of 50 μL
M
were used. The wavelength selected to detect MON was 227 nm.
2.3. Preparation of SAMs on gold electrodes
d
Different SAMs were prepared from immersion of the activated Au electrodes in
Ac ce pt e
stirred ethanolic solutions (200 rpm) of thiols in a concentration range from 1 x 10-3 to 18 x 10-3 mol L-1 during different modification times (tmod: 5 – 120 min). These experiments were carried out in darkness. Finally, modified electrodes were rinsed with ethanol and CBS to remove the physically adsorbed thiols.
2.4. Sample preparation
The maize samples to perform electrochemical and HPLC measurements were
prepared following a procedure similar to that described previously by Parich [12] with some modifications. Two sterile samples of maize were used. One of them was contaminated with a Fusarium temperatum strain and the other was free of the toxin to be
8
Page 8 of 34
used as negative control. Briefly, maize samples were milled, and MON was extracted with an acetonitrile/water (84:16) solution. The acetonitrile in the extract was evaporated on a rotary evaporator. The extracts were diluted with methanol, and stirred until a
ip t
homogeneous solution was obtained. At the same time, SAX Strong Anion Exchange column was conditioned by adding 2 mL of methanol followed by 2 mL of water and 2 mL
cr
of 0.1 mol L-1 phosphoric acid. Then, the extracts were introduced in SAX columns and
us
once the solution had passed, columns were washed with 0.1 mol L-1 phosphoric acid and water. Finally, MON was eluted as MON-tetrabutylammoniun ion pair from contaminated
an
maize samples. On the other hand, a MON free extract was obtained from negative control sample. Both extracts were used to prepare solutions to perform electrochemical and HPLC
3. Results and discussion
d
M
measurements.
Ac ce pt e
3.1. Characterization of the monolayer
Different thiols were used to generate the SAMs. These thiols were primary and
tertiary amines as well as aliphatic and aromatic such as CA, 2-DAET, 8-AOT, 4-ATP and 4-MP. However, the MON oxidation was only observed at electrodes modified with CA and 2-DAET. This result can be explained considering that CA and 2-DAET are shortchain thiols, giving rise to poorly ordered monolayers (see section 3.2). The characterization of SAMs was carried out by measurements of double layer
capacitance (Cdl). The Cdl is a sensitive measure of the state of the adsorbed monolayers on electrodes [33]. The Cdl was calculated from cyclic voltammograms registered in acid medium for both bare gold and Au-CA electrodes using the Eq. (1) [34].
9
Page 9 of 34
Cdl =
Ic vA
(1)
where Ic is the capacitive current, v is the scan rate and A is the electrode surface area.
ip t
The Cdl studies were carried out for several CA concentrations and modification
cr
times. Cdl decreases as both the CA concentration and the modification time increase. A Cdl of 16.5 μF cm-2 was determined for Au-CA electrodes, prepared using the optimum CA
us
concentration and modification time, which is smaller than the corresponding to the bare gold electrode (58.4 μF cm-2). These results are explained through the Smith and White
an
model, which considers a SAM as an irreversibly adsorbed monolayer such that all
M
acid/base head groups lie in a common plane which refers to as the plane of acid dissociation. On the other hand, it is also considered that the ions of the inert electrolyte
d
from the medium do not penetrate into the film [35]. Thus, when the electrode is modified,
Ac ce pt e
the ions layer on the surface of the electrode will be replaced by organic molecules (i.e., cysteamine) and the Cdl will decrease.
The Au-CA electrodes were also characterized by their reductive desorption in 0.5
mol L-1 KOH. Therefore, cyclic voltammograms recorded in the potential range from 0 to 1.2 V showed two reduction peaks centered at -0.70 and - 1.0 V. The occurrence of multidesorption peaks can be attributed to desorption of thiol molecules adsorbed on different crystallographic domains of the polycrystalline gold electrode. An average surface coverage of (5±1) x 10-10 mol cm-2 was obtained for three different CA SAMs modified electrodes, close to the coverage expected for an Au(111) electrode, i.e., 7.7 x 10-10 mol cm2
[21]. Therefore, both studies of Cdl, and the reductive desorption of Au-CA electrodes
10
Page 10 of 34
were performed under the optimal modification conditions, i.e., CA concentration = 2.5 x 10-3 mol L-1 and tmod = 30 min (see below).
ip t
3.2. Electrochemical oxidation of MON on the Au-CA electrode Cyclic voltammograms for both the blank and MON (2.4 x 10-4 mol L-1) in
cr
pH 4 CBS solution recorded using bare gold electrodes are shown in Fig. S1 of
us
Supplementary Material.. The MON oxidation peak was overlapped with the corresponding wave of formation of gold oxides. Previous studies of MON oxidation on glassy carbon
an
electrodes in acetonitrile showed that MON is oxidized at potentials about 1.1 V vs Ag+ [19]. In order to achieve less anodic potentials for MON oxidation, different thiols were
M
checked by forming SAMs on a gold electrode. Thiols used were CA, 2-DAET, 8-AOT, 4ATP, and 4-MP, at a concentration of 2 x 10-3 mol L-1. The SAMs formation was
d
performed with a tmod of 60 min. In pH 4 CBS solution, the amino group of most of the
Ac ce pt e
thiols and the nitrogen atom at the 4-MP are mainly protonated and positively charged. Cyclic voltammograms were recorded in a 2.4 x 10-4 mol L-1 MON solution using electrodes modified with the different thiols. However, MON oxidation was only observed for the electrodes modified with CA and 2-DAET SAMs. Therefore, a pre-concentration of MON was performed using an accumulation time (tacc) of 80 min. Voltammograms obtained on both Au-CA and Au-2-DAET in the absence and in the presence of MON are shown in Fig. 2. They show one oxidation peak in the anodic sweep, centred at 0.38 and 0.45 V vs Ag/AgCl for Au-CA and Au-2-DAET electrodes, respectively, which could be assigned to the oxidation of MON. These oxidation potentials are shifted 0.7 V negatively respect to the oxidation potential obtained under the same conditions at bare gold electrodes. It is assumed that attractive interactions between both CA-SAMs and 2-DAET11
Page 11 of 34
SAMs positively charged and MON negatively charged at pH 4 CBS is the main responsible to reduce the MON oxidation potential at modified electrodes compared with the MON oxidation potential on the bare gold electrode. A similar behaviour was found by
ip t
us when the same modified electrode (Au-CA) was used in the determination of ochratoxin A in red wine samples [24]. Anodic peak currents (Ip,a) were linear with v, showing an
cr
adsorption control for the electrode process [30]. The secondary peaks found at about 0.15
us
V and 0.25 V in both solutions, respectively, in the absence and in the presence of MON are considered to be caused by the CBS itself and do not interfere with the response of
an
MON. The cathodic sweep showed an increase in the capacitive currents, indicating the adsorption of species produced during the MON oxidation and a reduction peak close to 0
M
V vs Ag/AgCl for both modified electrodes, which could be associated to the reduction of a product of MON oxidation. The reduction peak currents were also linear with v, showing
d
an adsorption control for the electrode process. Moreover, if the forward scan is reversed at
Ac ce pt e
a potential of 0.2 V, the cathodic peak is not observed, clearly highlighting the cathodic peak is related to the oxidation peak of MON. In addition, it is well known that aldehydes and ketones react with primary amines to form imines (Schiff bases) [36]. The use of a wider range of thiols was previously explained. The different variables
studied
for
Au-CA
electrodes
were
also
studied
for
Au-2-DAET electrodes. In all cases, the best electrochemical responses were obtained at Au-CA electrodes. Therefore, they were used to construct the electrochemical sensor to quantify MON in maize samples. Thus, the oxidation of MON on cysteamine SAMs may be favoured by both electrostatic interactions and the formation of the Schiff base between MON and non-protonated cysteamine on the electrode surface, as a part of its acid/base equilibrium. As MON is oxidized at a lower anodic potential on Au-CA than on the Au-212
Page 12 of 34
DAET and the oxidation peak is better defined, CA was used for the development of the sensor. Therefore, parameters such as concentration of CA, modification and accumulation
ip t
times were adjusted to reach the maximum performance of the Au-CA sensor.
3.3. Determination of optimum cysteamine concentration and electrode modification
cr
time
us
The SAMs were prepared by immersion of clean gold electrodes in CA ethanolic solutions [21,24]. Two parameters were investigated, the concentration of thiol and the
an
modification time and their performance were evaluated thorough the Ip,a of MON oxidation. The concentration of CA was varied in a range between 0.5 and 8 x 10-3 mol L-1,
M
and the resulting SAMs were evaluated for a given concentration of MON and tacc = 60 min. The current obtained for MON oxidation increased with CA concentration, reaching a
d
steady value at a CA concentration of about 2.5 x 10-3 mol L-1, as shown in Fig. 3A. Thus,
Ac ce pt e
this CA concentration was used for all experiments. This concentration value is in good agreement with results of other studies, where the best conditions were achieved at low concentrations of CA [24,37].
The tmod is a very important parameter, which is related directly to the concentration
of thiol and defines the properties of monolayers (wettability, order and coverage). The tmod was evaluated between 5 and 120 min for given concentrations of CA (2.5 x 10-3 mol L-1) and MON (2.4 x 10-4 mol L-1). A maximum current was reached when the electrode was modified during 30 min, as shown in Fig. 3B. After this time, the current started to decrease. This behaviour can be explained by the formation of SAMs more compact. This result is in a good agreement with results of Wirde et al. [37], who reported that with only 5 min of tmod just 80 % of the electrode surface is covered by CA. Moreover, the permeation 13
Page 13 of 34
of the ions and redox couples into the monolayer are also affected by the order of the structure of SAMs [38,39]. The permeation into the monolayer will decrease when the structure is highly ordered. In this case, when the tmod is 30 min the structure of the
ip t
monolayer allows the maximum permeability of MON and other ions into it.
cr
3.4. Determination of MON accumulation time
us
A study for the determination of the tacc of MON on Au-CA was performed. Cyclic voltammograms were recorded at different tacc in a range from 5 to 160 min to establish the
an
time at which the maximum surface concentration of MON occurs. This study was carried out for MON concentrations in the range between 1 x 10-9 and 1 x 10-7 mol L-1. Plots of Ip,a
M
vs tacc at different concentrations of MON (Fig. 4) showed that the Ip,a increased with the tacc. Then, at about 80 min, a saturation of surface with MON was reached and constant
d
values of Ip,a were obtained. The same result was observed for peak potentials, where they
Ac ce pt e
become less anodic with increasing of tacc and reached a constant value at about 80 min (data not shown). This behaviour is expected from such kind of systems, wherein the electroactive species is accumulated on the electrode surface [40]. In this case, the accumulation is enhanced by both the electrostatic attraction between anionic MON and Au-CA positively charged and the probable formation of a Schiff base. Thus, all measurements were performed at a tacc = 80 min. On the other hand, it was found a linear relationship between anodic and cathodic
charges obtained from oxidation and reduction processes, respectively, for different tacc (Results no shown). Results of linear regression were: slope = 0.57 ± 0.002, intercept = (9.90 ± 0.42) x 10-7 and a linear correlation coefficient (r) of 0.9954.
14
Page 14 of 34
3.5. Calibration curve A calibration curve was constructed from CV results obtained from solutions
ip t
prepared from the commercial reagent. MON standard solutions were prepared in the concentration range from 1 x 10-9 to 1 x 10-7 mol L-1 in CBS, under optimized conditions
cr
previously described. Cyclic voltammograms recorded to construct the calibration curve are
us
shown in Fig. S2 of Supplementary Material. A linear relationship between the anodic peak current and the MON concentration was found. The points used to construct the calibration
an
curve were obtained by triplicate (six points were taken into account, r = 0.9863). The calibration curve is shown in Fig. S3 of Supplementary Material. It is expressed by a least-
M
square procedure as:
(2)
Ac ce pt e
d
I p ,a ( A ) = ( 7.59 ± 0.14 ) x10−7 ( A ) + (4.35 ± 0.28) (A mol L−1 ) c*MON (mol L−1 )
The LOD was 8.3 x 10-10 mol L-1 (0.1 ppb) for a signal-to-noise ratio of 3:1 [41].
The repeatability and reproducibility of the electrochemical sensor was checked
using standard solutions of MON of 2.49 x 10-9 and 2.49 x 10-8 mol L-1. The repeatability was tested by performing three consecutive measurements on the same solution. Values of RSD% were 3.1 % and 3.6 % for 0.3 and 3 ppb, respectively. On the other hand, the reproducibility was obtained from measurement accomplished in three consecutive days. Values obtained were 3.1 % and 3.0 % for 2.49 x 10-9 and 2.49 x 10-8 mol L-1, respectively. These results show that the electrochemical sensor to determine MON has a very good performance.
15
Page 15 of 34
3.6. Analytical determination of MON in maize samples 3.6.1. Determination of MON using the Au-CA electrochemical sensor
ip t
From the extracts obtained as described in Section 2.4, solutions were prepared with both, contaminated and uncontaminated extracts in pH 4 CBS at 10 % v/v, and the cyclic
cr
voltammograms were recorded using the Au-CA sensor (Fig. 5). The solution with the
us
contaminated extract showed an oxidation peak at about 0.45 V vs Ag/AgCl, which can be assigned to the MON oxidation present in the sample. This oxidation peak presented a
an
small potential shift of 0.07 V compared to the peak potential obtained using the commercial reagent. This shift may be caused by matrix effects itself, but we have not
M
found a significant effect in the analytical determination of MON. Regarding to the response obtained with the sensor in the solution with the negative control extract, a wave
d
at about 0.63 V vs Ag/AgCl was observed. This wave is located at higher potentials than
Ac ce pt e
those corresponding to the oxidation peak of MON, and might be assigned to the oxidation of interfering substances present in the matrix. Even when the extract was obtained from an extraction in solid phase using an ESP-SAX column, some components of maize sample could pass through the column and be oxidized in the range of potential used. To minimize the matrix effect, the standard addition method with the Au-CA sensor was used to determine MON in maize samples. A solution with 0.5 % of the contaminated extract in CBS was prepared. This solution was spiked with different amounts of standard solution of MON in order to achieve desired concentrations. The curve obtained for spiked sample was linear in the range from 4.15 x 10-7 to 1.24 x 10-6 mol L-1, with r = 0.9942 and can be represented by at least-square procedure as:
16
Page 16 of 34
(
)
(
I p,a ( A ) = (1.34 ± 0.04 ) x10−6 ( A ) + (1.20 ± 0.05 ) A L mol−1 c*MON mol L−1
)
(3)
10-4 mol L-1.
cr
3.6.2. Determination of MON by HPLC with UV-Vis detection
ip t
Thus, the concentration of MON obtained for the maize sample was (2.23 ± 0.06) x
us
To compare the result obtained from the determination of MON in maize samples using the Au-CA electrochemical sensor, a HPLC/UV-Vis methodology proposed by
an
Parich et al. [12] was performed. The MON concentration determined by HPLC was considered as the reference value.
M
A calibration curve (peak area vs c*MON) was constructed using standard solutions of MON dissolved in the same mobile phase. A linear relationship was found in the
d
concentration range from 4.16 x 10-7 to 1.00 x 10-5 mol L-1 with the following parameters:
Ac ce pt e
intercept = (0.03 ± 0.06), slope = (1.01 ± 0.02) x 106 mol-1 L and r = 0.9991. The LOD was 8.3 x 10-9 mol L-1 (1 ppb) for a 3:1 signal-to-noise ratio and the limit of quantification (LOQ) was 2.49 x 10-8 mol L-1 (3 ppb) for a 10:1 signal-to-noise ratio. On the other hand, a concentration of MON in the contaminated maize sample of (1.92 ± 0.03) x 10-4 mol L-1 was determined by HPLC/UV-Vis. Thus, the concentration value of MON in the contaminated maize sample obtained by both methodologies agrees satisfactorily, indicating the very good performance of the Au-CA electrochemical sensor.
17
Page 17 of 34
Conclusions This is the first report which describes the development of a very sensitive electrochemical sensor for the MON determination in maize samples based on the
ip t
electrochemical oxidation of MON adsorbed at cysteamine self assembled monolayers on gold electrodes. A very good limit of detection of 8.3 x 10-10 mol L-1 (0.1 ppb) and RDS%
cr
less than 4 % in all concentration range were found. The limit of detection obtained was
us
about 400-fold lower than that found using the HPLC method proposed by Parich et al. [12] and others reported in the literature, which shows the good analytical characteristics of the
an
sensor. Also, the measurement of the concentration of MON in maize samples using the sensor and the HPLC methodology were in very good agreement, showing the good
M
performance of the sensor. This Au-CA electrochemical sensor can provide a promising
Ac ce pt e
low cost equipment.
d
prospect for the determination of MON in other real samples using a simple method and a
Acknowledgment
The authors thank Dra. Sofia Chulze for providing real samples of maize and the
HPLC equipment. Financial support from the Consejo Nacional de Investigaciones Científicas y Técnicas (CONICET), Agencia Nacional de Promoción Científica y Tecnológica (FONCYT), Ministerio de Ciencia y Tecnología de la Provincia de Córdoba (MinCyT) and Secretaría de Ciencia y Técnica from the Universidad Nacional de Río Cuarto are gratefully acknowledged. P. C. Díaz Toro and M. V. Fumero thank CONICET for doctoral fellowships. We thank to Reviewers for their valuable suggestions.
18
Page 18 of 34
References [1] J. M. Soriano del Castillo, Micotoxinas en Alimentos, Díaz de Santos, España, 2007. [2] M. Jestoi, Emerging Fusarium – mycotoxins fusaproliferin, beauvericin, enniatins, and
ip t
moniliformin: a review, Crit. Rev. Food Sci. Nutr. 48 (2008) 21-49. [3] V. Scarpino, M. Blandino, M. Negre, A. Reyneri, F. Vanara, Moniliformin analysis in
cr
maize samples from North-West Italy using multifunctional clean-up columns and the
us
LC-MS/MS detection method, Food Addit. Contam. Part A 30 (2013) 876-884. [4] International Agency for Research on Cancer (IARC), 82, 2006.
an
[5] J. A. Engelhard, W. W. Carlton, J. F. Tuite, Toxicity of Fusarium moniliforme subglutinans for chicks, ducklings, and turkey poults, Avian. Dis. 33 (1989) 357-360.
M
[6] D. R. Ledoux, A. J. Bermudez, G. E. Rottinghaus, J. Broomhead, G. A. Bennett, Effects of feeding Fusarium fujikuroi culture material, containing known levels of
d
moniliformin, in young broiler chicks, Poult. Sci. 74 (1995) 297-305.
Ac ce pt e
[7] L. Z. Zhu, Y. M. Xia, G. Q. Yang, Blood glutathione peroxidase activities of populations in Keshan disease affected and non-affected areas, Acta. Nutr. Sin. 4 (1982) 229-233.
[8] P. G. Thiel, A molecular mechanism for the toxic action of moniliformin produced by Fusarium moniliforme, Biochem. Pharmacol. 27 (1978) 483-486. [9] L. T. Burka, J. Doran, B. J. Wilson, Enzyme inhibition and the toxic action of moniliformin and other vinylogous α-ketoacids, Biochem. Pharmacol. 31 (1982) 79-84. [10] M. C. Pirrung, S. K. Nauhaus, B. Singh, Cofactor-directed, time-dependent inhibition of thiamine enzymes by the fungal toxin moniliformin, J. Org. Chem. 61 (1996) 25922593.
19
Page 19 of 34
[11] N.W. Turner, S. Subrahmanyam, S.A. Piletsky, Analytical methods for determination of mycotoxins: A review, Anal Chim Acta 632(2009) 168-180. [12] A. Parich, L.S. Boeira, S.P. Castro, R. Krska, Determination of moniliformin using
ip t
SAX column clean-up and HPLC/DAD-detection, Mycotox. Res. 19(2003) 203-206. [13] C. Munimbazi, L. Bullerman, Chromatographic method for the determination of the
cr
mycotoxin moniliformin in corn, in: M. Trucksess, A. Pohland (Eds.), Mycotoxin
us
Protocols, Humana Press, 2001, pp. 131-145.
[14] G. Filek, W. Lindner, Determination of the mycotoxin moniliformin in cereals by
an
high-performance liquid chromatography and fluorescence detection, J Chromatogr A 732 (1996) 291-298.
Mycotox. Res. 14 (1998) 35-40.
M
[15] F. Schütt, H. Nirenberg, G. Demi, Moniliformin production in the genus Fusarium,
d
[16] M. Jestoi, M. Rokka, A. Rizzo, K. Peltonen, Moniliformin in Finnish grains: Analysis
Ac ce pt e
with LC-MS/MS, Aspect Appl. Biol. 68 (2003) 94-99. [17] C. M. Maragos, Detection of moniliformin in maize using capillary zone electrophoresis, Food Addit. Contam. 21 (2004) 803-810. [18] J. Gilbert, J. R. Startin, I. Parker, M. J. Shepherd, J. C. Mitchell, M. J. Perkins, Derivatization of the Fusarium mycotoxin moniliformin for gas chromatography-mass spectrometry analysis, J. Chromatogr. A 369 (1986) 408-414. [19] P. C. Díaz Toro, F. J. Arévalo, M. A. Zon, H. Fernández, Studies of the electrochemical behavior of moniliformin mycotoxin and its sensitive determination at pretreated glassy carbon electrodes in a non-aqueous medium, J Electroanal. Chem. 738 (2015) 40-46.
20
Page 20 of 34
[20] J. C. Love, L. A. Estroff, J. K. Kriebel, R. G. Nuzzo, G. M. Whitesides, Selfassembled monolayers of thiolates on metals as a form of nanotechnology, Chem. Rev. 105 (2005) 1103-1170.
ip t
[21] H. O. Finklea, Electrochemistry of organized monolayers of thiols and related
Chemistry, Marcel Dekker, New York, 1996, pp. 109-335.
cr
molecules on electrodes, in: A.J. Bard, I. Rubinstein (Eds.), Electroanalytical
Electrochem. Commun. 9 (2007) 2827-2832.
us
[22] M. Sheffer, V. Vivier, D. Mandler, Self-assembled monolayers on Au microelectrodes,
an
[23] H. O. Finklea, S. Avery, M. Lynch, T. Furtsch, Blocking oriented monolayers of alkylmercaptans on gold electrodes, Langmuir 3 (1987) 409-413.
M
[24] P. R. Perrotta, N. R. Vettorazzi, F. J. Arévalo, A. M. Granero, S. N. Chulze, M. A. Zon, H. Fernández, Electrochemical studies of ochratoxin A mycotoxin at gold
d
electrodes modified with cysteamine self-assembled monolayers. Its ultra sensitive
Ac ce pt e
quantification in red wine samples, Electroanalysis 23 (2011) 1585-1592. [25] M. Wanunu, A. Vaskevich, S. R. Cohen, H. Cohen, R. Arad-Yellin, A. Shanzer, I. Rubinstein, Branched coordination multilayers on gold, J. Am. Chem. Soc. 127 (2005) 17877-1787.
[26] S. V. Pereira, F. A. Bertolino, G. A. Messina, J. Raba, Microfluidic immunosensor with gold nanoparticle platform for the determination of immunoglobulin G antiEchinococcus granulosus antibodies, Anal. Biochem. 409 (2011) 98-104. [27] R. K. Shervedani, M. Bagherzadeh, S. A. Mozaffari, Determination of dopamine in the presence of high concentration of ascorbic acid by using gold cysteamine selfassembled monolayers as a nanosensor, Sens. Actuators B 115 (2006) 614-621.
21
Page 21 of 34
[28] M. V. Fumero, M. M. Reynoso, S. Chulze, Fusarium temperatum and Fusarium subglutinans isolated from maize in Argentina, Int. J. Food Microbiol. 199 (2015) 8692.
ip t
[29] D. T. Sawyer, J. L. Roberts, Experimental Electrochemistry for Chemists, 2nd Ed., New York: Wiley, 1995.
cr
[30] J. O. M. Bockris, B. E. Conway, R. E. White, Modern Aspects of Electrochemistry:
us
Springer, 1992.
[31] E. F. Douglass Jr, P. F. Driscoll, D. Liu, N. A. Burnham, C. R. Lambert, W. G. Mc
an
Gimpsey, Effect of electrode roughness on the capacitive behavior of self-assembled monolayers, Anal. Chem. 80 (2008) 7670-7677.
M
[32] J. C. Hoogvliet, M. Dijksma, B. Kamp, W. P. van Bennekom, Electrochemical Pretreatment of polycrystalline gold electrodes to produce a reproducible surface
Ac ce pt e
(2000) 2016-2021.
d
roughness for self-assembly: A study in phosphate buffer pH 7.4, Anal. Chem. 72
[33] B. B. Damaskin, O. A. Petrii, V. V. Batrakov, Adsorption of organic compounds on electrodes, Plenum Press, New York, USA, 1971.
[34] Z. Q. Feng, S. Imabayashi, T. Kakiuchi, K. Niki, Electroreflectance spectroscopic study of the electron transfer rate of cytochrome c electrostatically immobilized on the ω-carboxylalkanethiol monolayer modified gold electrode, J. Electroanal. Chem. 394 (1995) 149-154.
[35] C. P. Smith, H. S. White, Voltammetry of molecular films containing acid/base groups, Langmuir, 9(1993) 1-3. [36] T. W. Solomons, Fundamentals of Organic Chemistry, J. Wiley & Sons, New York, USA, 1982. 22
Page 22 of 34
[37] M. Wirde, U. Gelius, Self-assembled monolayers of cystamine and cysteamine on gold studied by XPS and voltammetry, Langmuir 15 (1999) 6370-6378. [38] M. J. Esplandiú, H. Hagenström, D. M. Kolb, Functionalized self-assembled
ip t
alkanethiol monolayers on Au(111) electrodes: 1. Surface structure and electrochemistry, Langmuir 17 (2001) 828-838.
cr
[39] J. J. Calvente, G. López-Pérez, P. Ramírez, H. Fernández, M. A. Zon, W. H. Mulder,
us
R. Andreu, Experimental study of the interplay between long-range electron transfer and redox probe permeation at self-assembled monolayers: Evidence for potential-
an
induced ion gating, J. Am. Chem. Soc. 127 (2005) 6476-6486.
[40] J. C. Love, D. B. Wolfe, R. Haasch, M. L. Chabinyc, K. E. Paul, G. M. Whitesides, R.
M
G. Nuzzo, Formation and structure of self-assembled monolayers of alkanethiolates on palladium, J. Am. Chem. Soc. 125 (2003) 2597-2609.
d
[41] D. MacDougall, W. B. Crummett, Guidelines for data acquisition and data quality
Ac ce pt e
evaluation in environmental chemistry, Anal. Chem. 52 (1980) 2242-2249.
23
Page 23 of 34
Figure Captions Figure 1. Chemical structure of moniliformin. X = Na, K or H.
ip t
Figure 2. Cyclic voltammograms recorded in pH 4 CBS at Au-CA (A) and Au-2-DAET (B) modified gold electrodes in blank (1) and in MON (2) solutions. c*MON = 2.4 x 10-4 mol
us
cr
L-1. cthiol = 2 x 10-3 mol L-1. tmod= 60 min. tacc = 80 min. v = 0.100 V s-1.
Figure 3. Variation of the anodic peak current with: (A) the concentration of cysteamine
an
for a modification time of 60 min, and (B) the modification time at a cysteamine concentration of 2.5 x 10-3 mol L-1 (B). In both experiments the concentration of MON was
M
2.4 x 10-4 mol L-1. The dash line was plotted to show the trend.
d
Figure 4. Variation of the anodic peak current with the accumulation time (tacc) at three
Ac ce pt e
concentration levels of MON. c*MON = 1 x 10-9 (1), 5 x 10-8 (2), 1 x 10-7 mol L-1 (3). The dash line was plotted to show the trend.
Figure 5. Cyclic voltammograms recorded with the Au-CA sensor in a solution with the negative control (1) and the contaminated extract (2). Both solutions were prepared in 10 % v/v pH 4 CBS. v = 0.100 V s-1.
Figure S1. Cyclic voltammograms of the blank (1) and MON in pH 4 CBS (2) solutions. c*MON = 2.4 x 10-4 mol L-1. v = 0.100 V s-1. Working electrode: bare gold electrode.
24
Page 24 of 34
Figure S2. Cyclic voltammograms recorded to construct the calibration curve using the Au-CA sensor in solutions of MON + pH 4 CBS. MON concentrations were: 1) 1.00 x 10-9;
Ac ce pt e
d
M
an
us
cr
Figure S3. The calibration curve obtained with data from Fig. S2.
ip t
2) 2.50 x 10-9; 3) 3.33 x 10-8; 4) 5.00 x 10-8 and 5) 1.00 x 10-7 mol L-1. v = 0.100 V s-1.
25
Page 25 of 34
Vitae Paulo César Díaz Toro obtained his Ph. D. in Chemistry (2015) from Río Cuarto National University (Río Cuarto, Argentina). He obtained a fellowship from Argentine Research
ip t
Council (CONICET) and was an active member of the Electroanalytical Group at the Chemistry Department, Faculty of Exact, Physico-Chemical and Natural Sciences (Río
cr
Cuarto National University). His research interest focuses on the development of
us
electroanalytical techniques for the determination of mycotoxins as well as the design and
an
characterization of electrochemical sensors.
Fernando J. Arévalo obtained his Ph. D. in Chemistry (2009) from Río Cuarto National
M
University (Río Cuarto, Argentina). He is a Researcher at Argentine Research Council (CONICET). At present, he also is Assistant Professor at the Chemistry Department,
d
Faculty of Exact, Physico-Chemical and Natural Sciences (Río Cuarto National
Ac ce pt e
University). Dr. Arévalo is an active member of the Electroanalytical Group at the Chemistry Department, and his research interest focuses on the development and characterization of electrochemical (bio) sensors based on the use of nano-structured materials.
María Verónica Fumero obtained her graduate in Microbiology (2012) from Río Cuarto National University (Río Cuarto, Argentina). She is actually doing a Ph. D. in Biological Sciences in the Microbiology and Immunology Department, Faculty of Exact, PhysicoChemical and Natural Sciences (Río Cuarto National University). At present, she has a doctoral fellowship from Argentine Research Council (CONICET) at the same Department. She is an active member of the Mycology and Mycotoxicology Group at the Microbiology 26
Page 26 of 34
and Immunology Department, and her research interest focuses on the study of species of Fusarium pathogens of maize and producers of diverse mycotoxins.
ip t
María A. Zon obtained her Ph. D. in Chemistry (1985) from Río Cuarto National University (Río Cuarto, Argentina). She did the postdoctoral training at Cordoba University
cr
(Córdoba, España) between 1990 and 1992. She is Full Professor at the Río Cuarto
us
National University and Principal Researcher at Argentine Research Council (CONICET). She has been the secretary of the Analytical Chemist Argentine Association (2007-2009).
an
Her research interest focuses on the development of electrochemical (bio) sensors by using nano-materials for the determination of different substrates such as mycotoxins, natural
M
antioxidants and hormones. She has over 65 peer-reviewed papers and eight book chapters. She is co-author of a book. She has been co-editor of an electroanalytical book. Prof. Zon is
Ac ce pt e
d
an AAQA, AAIFQ and SIBAE fellow.
Héctor Fernández obtained his Ph. D. in Chemistry (1978) from Río Cuarto National University (UNRC) (Río Cuarto, Argentina). He did the postdoctoral training (1980-1982) at the University of New York at Buffalo, Buffalo (USA). Currently, he is Full Professor at UNRC and Principal Researcher at Argentine Research Council (CONICET). He was Dean of the Faculty of Exact, Physico-Chemical and Natural Sciences (UNRC, 1992-1999) and Head of the Department of Chemistry at the Faculty of Exact, Physico-Chemical and Natural Sciences (2001-2004). He was President of the Argentinean Society of Analytical Chemists (2007-2009). His research interest focuses on several subjects, such as electrochemistry of mycotoxins, hormones and synthetic and natural antioxidants, studies on ultramicroelectrodes and electrodes modified by self-assembled monolayers of thiols, 27
Page 27 of 34
carbon nanotubes, antibodies, etc and their use for electroanalytical applications. Development of electroanalytical techniques for the determination of antioxidants, mycotoxins and hormones in real matrixes (plants, cereal, foods, sera of animal origin, etc,
ip t
respectively). Design and characterization of chemical sensors, electrochemical (bio)sensors and immunoelectrodes based on nanostructured materials. He has over eighty
cr
peer-reviewed papers, eight book chapters, co-author of a book and has been the editor of a
us
book. Prof. Fernández belongs to the Editorial Board of J Biosensors and Bioelectronics and Polish Journal of Environmental Studies. He is an AAQA, AAIFQ, SIBAE and ISE
An electrochemical sensor was developed to determine moniliformin in maize
•
Ac ce pt e
samples.
d
•
M
an
fellow.
A self-assembled monolayer modified gold electrode was used as working electrode.
• • •
Cyclic voltammetry was the electrochemical technique used. The calibration curve was linear in the concentration range from 0.12 to 12 ppb. Results of the electroanalytical method were in good agreement with those of HPLC.
28
Page 28 of 34
ed pt ce Ac
Page 29 of 34
a M ed pt ce Ac
Page 30 of 34
M a d te ep Ac c
Page 31 of 34
us an M ed pt ce Ac
Page 32 of 34
Page 33 of 34
ed
ce pt
Ac M an
c
us
ed pt ce Ac Page 34 of 34
S0925-4005(15)30623-7 http://dx.doi.org/doi:10.1016/j.snb.2015.11.042 SNB 19300
To appear in:
Sensors and Actuators B
Received date: Revised date: Accepted date:
9-6-2015 15-10-2015 7-11-2015
Please cite this article as: P.C.D. Toro, F.J. Ar´evalo, M.V. Fumero, M.A. Zon, H. Fern´andez, Very sensitive electrochemical sensor for moniliformin detection in maize samples, Sensors and Actuators B: Chemical (2015), http://dx.doi.org/10.1016/j.snb.2015.11.042 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ip t
Very sensitive electrochemical sensor for moniliformin detection in maize samples
Paulo César Díaz Toroa, Fernando Javier Arévaloa, María Verónica Fumerob, María Alicia Zona*
Departamento de Química. Universidad Nacional de Río Cuarto. Agencia Postal N° 3.
an
a
us
cr
and Héctor Fernándeza*
(X5804ALH) - Río Cuarto. Fax: 54-358-4676233; Tel: 54-358-4676233; Argentina. Departamento de Microbiología e Inmunología. Universidad Nacional de Río Cuarto.
M
b
Agencia Postal N° 3. (X5804ALH) - Río Cuarto. Fax: 54-358-4676231; Tel: 54-358-
Ac ce pt e
d
4676231. Argentina.
* Correspondig Authors
E-mail addresses: [email protected] (P. C. Díaz Toro); [email protected] (F. J. Arévalo); [email protected] (M. V. Fumero); [email protected] (M. A. Zon); [email protected] (H. Fernández).
1
Page 1 of 34
Abstract A highly sensitive electrochemical sensor based on self assembled monolayers (SAMs) of cysteamine on gold electrodes (Au-CA) in pH 4 citrate buffer solutions was
ip t
developed to quantify the moniliformin (MON) mycotoxin by cyclic voltammetry in contaminated maize samples. Parameters of SAMs modified gold electrodes such as
cr
concentration of thiol, modification time and MON accumulation time were adjusted to
us
obtain the best performance of the sensor. The calibration curve was linear in the MON concentration range from 1 x 10-9 to 1 x 10-7 mol L-1. A limit of detection of 8.3 x 10-10 mol
an
L-1 (0.1 ppb) was obtained. The standard addition method was used to minimize matrix effects. The concentration of MON determined with the electrochemical sensor was in very
M
good agreement with results found by HPLC UV-Vis. The sensor showed a good analytical
Ac ce pt e
d
performance with a relative standard deviation (RDS%) of about 3%.
Keywords: Mycotoxins; Moniliformin; Self-assembled monolayers; Cyclic voltammetry; Maize samples.
2
Page 2 of 34
1. Introduction
Moniliformin (3-hydroxy-3-cyclobutene-1,2-dione, MON, Fig. 1) is a mycotoxin produced by different species of Fusarium, mainly F. proliferatum [1]. In the nature, MON
ip t
occurs as the sodium or potassium salt, and presents a pKa less than 1, which makes it soluble in polar solvents [1].
cr
The mechanism of toxicity of MON is not yet well known, and its toxic-kinetics is
us
still unknown. MON was found in cereals such as maize, rice, barley, wheat, oats, rye and triticale [2], although the greatest contamination was found in maize samples [3]. The
an
International Agency for Researcher on Cancer (IARC) classifies MON in the Group 3, because the agent is not classifiable as carcinogenic to humans [4]. Several studies have
M
confirmed the toxic effect produced by MON in animals [5,6] and its potential responsibility in the Keshan disease that occurred in China [7]. It has been proposed that
d
MON inhibits the oxidation of intermediates in the tricarboxylic acid cycle [8,9], causing
Ac ce pt e
respiratory diseases, and myocardial degeneration in animals, and leading, in some cases, death. It is also proposed that MON competes with pyruvate for active sites of α– ketoglutarate dehydrogenase, pyruvate decarboxylase, and aceto-hydroxy acid synthetase enzymes [10]. Thus, MON was classified into a group of emerging mycotoxins, highlighting its potential toxic effects on both animals and humans [11]. Until now, there is no regulations respect the maximum levels of MON allowed in food or grains, or even an official technique for its quantification [2, 11]. Thus, the development of relatively fast analytical methods to determine MON in contaminated samples is a major challenge, where sensitivity, selectivity, precision, and accuracy are required. However, several methods have been developed to quantify MON based mostly on chromatography techniques,
3
Page 3 of 34
especially the high performance liquid chromatography (HPLC). All these techniques require a pre-treatment of the samples to achieve a good performance. Often, derivatization reactions or ion-pair agents are necessary to obtain products that can be detected by HPLC
ip t
UV-Vis [12,13], HPLC fluorescence [14], TLC-UV [15], LC-MS/MS [16]. The technique of capillary zone electrophoresis with diode array detection (CZE-DAD) uses a similar pre-
cr
treatment to that of HPLC [17]. The limits of detection (LOD) of these techniques are
us
higher than of 40 ppb. Gilbert et al [18] have developed a methodology to detect MON by GC-MS and found a LOD of 0.5 ppb. The major problem of this method is the
an
derivatization step of MON with N-methyl-N-(tert-butyldimethylsilyl) trifluroacetamide containing 1 % tert-butyldimethylchlorosilane. We have recently determined the
M
thermodynamic and kinetics parameters of MON oxidation at glassy carbon electrodes in acetonitrile solution [19]. We have also proposed an electroanalytical methodology to
Ac ce pt e
satisfactory.
d
detect MON in this medium, but the determination of MON in maize samples was not
The use of self assembled monolayers (SAMs) to functionalize clean metallic
surfaces (either obtained by soaking the surfaces in solutions or in vapour phase) has been widely used in the development of electroanalytical techniques to quantify of a large number of analytes [20]. These SAMs are characterized by their high degree of orientation and stability due to the thiol is covalently bonded to the corresponding metallic substrate, particularly, gold, platinum, and mercury [21]. In addition, different interaction forces are present among adsorbed molecules, such as Van der Waals forces, hydrogen bond, and πinteractions [21]. Organosulfur compounds can be aliphatic or aromatic, and have different chain length and variable tail functional groups. Tail groups can be chosen to achieve a better affinity and specificity towards a given analyte, and provide selectivity and 4
Page 4 of 34
sensitivity to the sensor. However, the final response depends on the alkane-thiol chain length and/or the nature of tail functional groups [21,22]. Short alkane-thiol SAMs form thin monolayers with high defect sites generating pinholes [23], where the electrocatalytic
ip t
activity of different substrates has been observed [24]. The major advantages of SAMs with respect to other surface modification is the well defined control over the composition,
cr
structure, thickness and orientation of the monolayer, which is given by the nature and the
us
alkane-thiol chain length [25,26]. On the other hand, monolayers from organosulfur compounds spontaneously self-assembled onto clean metallic surfaces are easy and simple
an
to form and require not special equipments [21].
Because of its low pKa, MON is present in its deprotonated form in pH 4 buffer
M
solutions, while cysteamine (pKa ≅ 7.6) adsorbed on the gold electrode is mainly protonated at pH 4 [27]. Thus, in principle, the electrostatic interaction between MON and the gold
d
electrode modified with cysteamine SAMs can be used to pre-concentrate MON, and
Ac ce pt e
electro-catalyze its electrochemical oxidation.
In this work, we describe the development of a very sensitive electrochemical
sensor to determine MON in maize samples, which is based on MON oxidation on cysteamine SAMs modified polycrystalline gold electrodes (Au-CA) in pH 4 citrate buffer solutions (CBS). The electrochemical technique used was cyclic voltammetry (CV). The results obtained by the electrochemical method were compared with those obtained by HPLC, indicating a very good performance of the proposed electrochemical sensor. As far as we know, this is the first electrochemical sensor reported for the detection of MON in real samples.
5
Page 5 of 34
2. Materials and methods 2.1. Reagents All reagents were of analytical grade. Sodium salt of MON, 2-aminoethanethiol
ip t
(cysteamine, CA), 2-(diethylamino) ethanethiol hydrochloride (2-DAET), 8-amino-1octanethiol hydrochloride (8-AOT), 4-amino-thiophenol (4-ATP) and 4-mercaptopyridine
cr
(4-MP) were purchased from Sigma-Aldrich (St. Louis, USA).CBS, ethanol, potassium
us
dihydrogen phosphate (KH2PO4), tetrabutylammonium hydroxide (TBAOH), phosphoric acid, sulfuric acid and hydrogen peroxide were purchased from Merck p.a. (Darmstadt,
Argentina). All reagents were used as received.
an
Germany). Acetonitrile and methanol were HPLC grade (Sintorgan, Buenos Aires,
M
A MON stock solution (4.2 x 10-3 mol L-1) was prepared in H2O (Mili-Q) and stored at 4ºC. Working solutions were prepared daily by adding aliquots of the stock solution to
d
CBS. The final concentration of MON was controlled by UV-Vis spectroscopy (see below).
Ac ce pt e
Solutions of thiols were prepared daily in ethanol at different concentrations and stored in darkness.
The mobile phase used for HPLC experiments was an ion pair buffer (10 mL of ion
pair solution and 50 mL of acetonitrile were diluted with water to 1 L). The ion pair solution was prepared by mixing 50 mL of 20 % (w/w) TBAOH in water and 100 mL of 1.1 mol L-1 KH2PO4. The pH was adjusted to 7.0 by addition of phosphoric acid solution (Sigma-Aldrich, St. Louis, USA).
Both uncontaminated and contaminated with Fusarium temperatum RCFT914 strain [28] maize samples were provided by the Departamento de Microbiología e Inmunología, Facultad de Ciencias Exactas, Físico-Químicas y Naturales, Universidad Nacional de Río Cuarto. 6
Page 6 of 34
2.2. Materials and apparatus Cyclic voltammetry (CV) measurements were carried out in a conventional threeelectrode cell. The working electrode was a polycrystalline gold disk (CH Instrument,
ip t
Austin, USA) of 2 mm of diameter. Previous to perform the experiments, the gold electrode was successively polished on BAS cloth with wet alumina powder (0.3 and 0.05 μm from
cr
Fischer, respectively), washed with water and cleaned in a H2SO4 + 30% H2O2 (3:1 v/v)
us
solution during 3 min. Then, it was placed in an ultrasonic bath for 5 min. Finally, it was cycled in 0.5 mol L-1 H2SO4 between 0.2 and 1.7 V until a typical voltammogram of a
an
polycrystalline Au clean surface was obtained [29]. The electrochemical area of the gold electrode was calculated through the charge of the cathodic peak for the reduction of their
M
oxides assuming a charge density of 420 μC cm-2, which corresponds to the reduction of one monolayer of gold oxide for polycrystalline gold [30]. A roughness factor of 5.4 ± 0.7
d
was calculated from the ratio of electrochemical and geometric surface areas and it was
Ac ce pt e
kept constant in all experiments [27]. A similar value of roughness factor was also found by other authors, who used a similar extensive pre-treatment prior to the modification of electrodes [31,32]. An Ag/AgCl (BAS, West Lafayette, USA) electrode and a Pt wire of large area (A ≈ 2 cm2) were used as the reference and auxiliary electrodes, respectively. Cyclic voltammograms were performed with a Micro AutoLab PGSTAT 101 potentiostat (Eco-Chemie, Utrecht, The Netherlands) controlled by the NOVA 1.10.2 software. All measurements were carried out at 25.0 ± 0.2 ºC. UV–Vis absorption spectra were recorded immediately after the preparation of MON solutions to determine their concentrations. A spectrophotometer Hewlett–Packard model 8452A was used. The UV-Vis measurements were performed using silica cells of 1 cm path length. Absorption spectra of MON were recorded at different concentrations in 7
Page 7 of 34
CBS. MON absorption spectra showed two bands, with maxima at λ = 227 nm and at λ = 258 nm, respectively. Plots of A vs c*MON at λ = 227 and λ = 258 nm were linear from 8.3 x
103 mol-1 L cm-1 and ε258 = (8.02 ± 0.05) x 103 mol-1 L cm-1.
ip t
10-6 to 4.5 x 10-5 mol L-1. Molar extinction coefficients in CBS were ε227 = (25.46 ± 0.05) x
A SAX Strong Anion Exchange column was used for clean-up of maize samples
cr
before CV and HPLC measurements. The HPLC measurements were performed with a
us
Hewlett Packard HPLC Serie 1100, coupled to UV-Vis detector. The column used was a Hypersil C18 (150 x 4.2 mm, 5 μm; Phenomenex) and a guard column ODS-Hypersil (20 x
an
2.1 mm, 5 μm; Phenomenex). A flow rate of 1 mL min-1 and an injection volume of 50 μL
M
were used. The wavelength selected to detect MON was 227 nm.
2.3. Preparation of SAMs on gold electrodes
d
Different SAMs were prepared from immersion of the activated Au electrodes in
Ac ce pt e
stirred ethanolic solutions (200 rpm) of thiols in a concentration range from 1 x 10-3 to 18 x 10-3 mol L-1 during different modification times (tmod: 5 – 120 min). These experiments were carried out in darkness. Finally, modified electrodes were rinsed with ethanol and CBS to remove the physically adsorbed thiols.
2.4. Sample preparation
The maize samples to perform electrochemical and HPLC measurements were
prepared following a procedure similar to that described previously by Parich [12] with some modifications. Two sterile samples of maize were used. One of them was contaminated with a Fusarium temperatum strain and the other was free of the toxin to be
8
Page 8 of 34
used as negative control. Briefly, maize samples were milled, and MON was extracted with an acetonitrile/water (84:16) solution. The acetonitrile in the extract was evaporated on a rotary evaporator. The extracts were diluted with methanol, and stirred until a
ip t
homogeneous solution was obtained. At the same time, SAX Strong Anion Exchange column was conditioned by adding 2 mL of methanol followed by 2 mL of water and 2 mL
cr
of 0.1 mol L-1 phosphoric acid. Then, the extracts were introduced in SAX columns and
us
once the solution had passed, columns were washed with 0.1 mol L-1 phosphoric acid and water. Finally, MON was eluted as MON-tetrabutylammoniun ion pair from contaminated
an
maize samples. On the other hand, a MON free extract was obtained from negative control sample. Both extracts were used to prepare solutions to perform electrochemical and HPLC
3. Results and discussion
d
M
measurements.
Ac ce pt e
3.1. Characterization of the monolayer
Different thiols were used to generate the SAMs. These thiols were primary and
tertiary amines as well as aliphatic and aromatic such as CA, 2-DAET, 8-AOT, 4-ATP and 4-MP. However, the MON oxidation was only observed at electrodes modified with CA and 2-DAET. This result can be explained considering that CA and 2-DAET are shortchain thiols, giving rise to poorly ordered monolayers (see section 3.2). The characterization of SAMs was carried out by measurements of double layer
capacitance (Cdl). The Cdl is a sensitive measure of the state of the adsorbed monolayers on electrodes [33]. The Cdl was calculated from cyclic voltammograms registered in acid medium for both bare gold and Au-CA electrodes using the Eq. (1) [34].
9
Page 9 of 34
Cdl =
Ic vA
(1)
where Ic is the capacitive current, v is the scan rate and A is the electrode surface area.
ip t
The Cdl studies were carried out for several CA concentrations and modification
cr
times. Cdl decreases as both the CA concentration and the modification time increase. A Cdl of 16.5 μF cm-2 was determined for Au-CA electrodes, prepared using the optimum CA
us
concentration and modification time, which is smaller than the corresponding to the bare gold electrode (58.4 μF cm-2). These results are explained through the Smith and White
an
model, which considers a SAM as an irreversibly adsorbed monolayer such that all
M
acid/base head groups lie in a common plane which refers to as the plane of acid dissociation. On the other hand, it is also considered that the ions of the inert electrolyte
d
from the medium do not penetrate into the film [35]. Thus, when the electrode is modified,
Ac ce pt e
the ions layer on the surface of the electrode will be replaced by organic molecules (i.e., cysteamine) and the Cdl will decrease.
The Au-CA electrodes were also characterized by their reductive desorption in 0.5
mol L-1 KOH. Therefore, cyclic voltammograms recorded in the potential range from 0 to 1.2 V showed two reduction peaks centered at -0.70 and - 1.0 V. The occurrence of multidesorption peaks can be attributed to desorption of thiol molecules adsorbed on different crystallographic domains of the polycrystalline gold electrode. An average surface coverage of (5±1) x 10-10 mol cm-2 was obtained for three different CA SAMs modified electrodes, close to the coverage expected for an Au(111) electrode, i.e., 7.7 x 10-10 mol cm2
[21]. Therefore, both studies of Cdl, and the reductive desorption of Au-CA electrodes
10
Page 10 of 34
were performed under the optimal modification conditions, i.e., CA concentration = 2.5 x 10-3 mol L-1 and tmod = 30 min (see below).
ip t
3.2. Electrochemical oxidation of MON on the Au-CA electrode Cyclic voltammograms for both the blank and MON (2.4 x 10-4 mol L-1) in
cr
pH 4 CBS solution recorded using bare gold electrodes are shown in Fig. S1 of
us
Supplementary Material.. The MON oxidation peak was overlapped with the corresponding wave of formation of gold oxides. Previous studies of MON oxidation on glassy carbon
an
electrodes in acetonitrile showed that MON is oxidized at potentials about 1.1 V vs Ag+ [19]. In order to achieve less anodic potentials for MON oxidation, different thiols were
M
checked by forming SAMs on a gold electrode. Thiols used were CA, 2-DAET, 8-AOT, 4ATP, and 4-MP, at a concentration of 2 x 10-3 mol L-1. The SAMs formation was
d
performed with a tmod of 60 min. In pH 4 CBS solution, the amino group of most of the
Ac ce pt e
thiols and the nitrogen atom at the 4-MP are mainly protonated and positively charged. Cyclic voltammograms were recorded in a 2.4 x 10-4 mol L-1 MON solution using electrodes modified with the different thiols. However, MON oxidation was only observed for the electrodes modified with CA and 2-DAET SAMs. Therefore, a pre-concentration of MON was performed using an accumulation time (tacc) of 80 min. Voltammograms obtained on both Au-CA and Au-2-DAET in the absence and in the presence of MON are shown in Fig. 2. They show one oxidation peak in the anodic sweep, centred at 0.38 and 0.45 V vs Ag/AgCl for Au-CA and Au-2-DAET electrodes, respectively, which could be assigned to the oxidation of MON. These oxidation potentials are shifted 0.7 V negatively respect to the oxidation potential obtained under the same conditions at bare gold electrodes. It is assumed that attractive interactions between both CA-SAMs and 2-DAET11
Page 11 of 34
SAMs positively charged and MON negatively charged at pH 4 CBS is the main responsible to reduce the MON oxidation potential at modified electrodes compared with the MON oxidation potential on the bare gold electrode. A similar behaviour was found by
ip t
us when the same modified electrode (Au-CA) was used in the determination of ochratoxin A in red wine samples [24]. Anodic peak currents (Ip,a) were linear with v, showing an
cr
adsorption control for the electrode process [30]. The secondary peaks found at about 0.15
us
V and 0.25 V in both solutions, respectively, in the absence and in the presence of MON are considered to be caused by the CBS itself and do not interfere with the response of
an
MON. The cathodic sweep showed an increase in the capacitive currents, indicating the adsorption of species produced during the MON oxidation and a reduction peak close to 0
M
V vs Ag/AgCl for both modified electrodes, which could be associated to the reduction of a product of MON oxidation. The reduction peak currents were also linear with v, showing
d
an adsorption control for the electrode process. Moreover, if the forward scan is reversed at
Ac ce pt e
a potential of 0.2 V, the cathodic peak is not observed, clearly highlighting the cathodic peak is related to the oxidation peak of MON. In addition, it is well known that aldehydes and ketones react with primary amines to form imines (Schiff bases) [36]. The use of a wider range of thiols was previously explained. The different variables
studied
for
Au-CA
electrodes
were
also
studied
for
Au-2-DAET electrodes. In all cases, the best electrochemical responses were obtained at Au-CA electrodes. Therefore, they were used to construct the electrochemical sensor to quantify MON in maize samples. Thus, the oxidation of MON on cysteamine SAMs may be favoured by both electrostatic interactions and the formation of the Schiff base between MON and non-protonated cysteamine on the electrode surface, as a part of its acid/base equilibrium. As MON is oxidized at a lower anodic potential on Au-CA than on the Au-212
Page 12 of 34
DAET and the oxidation peak is better defined, CA was used for the development of the sensor. Therefore, parameters such as concentration of CA, modification and accumulation
ip t
times were adjusted to reach the maximum performance of the Au-CA sensor.
3.3. Determination of optimum cysteamine concentration and electrode modification
cr
time
us
The SAMs were prepared by immersion of clean gold electrodes in CA ethanolic solutions [21,24]. Two parameters were investigated, the concentration of thiol and the
an
modification time and their performance were evaluated thorough the Ip,a of MON oxidation. The concentration of CA was varied in a range between 0.5 and 8 x 10-3 mol L-1,
M
and the resulting SAMs were evaluated for a given concentration of MON and tacc = 60 min. The current obtained for MON oxidation increased with CA concentration, reaching a
d
steady value at a CA concentration of about 2.5 x 10-3 mol L-1, as shown in Fig. 3A. Thus,
Ac ce pt e
this CA concentration was used for all experiments. This concentration value is in good agreement with results of other studies, where the best conditions were achieved at low concentrations of CA [24,37].
The tmod is a very important parameter, which is related directly to the concentration
of thiol and defines the properties of monolayers (wettability, order and coverage). The tmod was evaluated between 5 and 120 min for given concentrations of CA (2.5 x 10-3 mol L-1) and MON (2.4 x 10-4 mol L-1). A maximum current was reached when the electrode was modified during 30 min, as shown in Fig. 3B. After this time, the current started to decrease. This behaviour can be explained by the formation of SAMs more compact. This result is in a good agreement with results of Wirde et al. [37], who reported that with only 5 min of tmod just 80 % of the electrode surface is covered by CA. Moreover, the permeation 13
Page 13 of 34
of the ions and redox couples into the monolayer are also affected by the order of the structure of SAMs [38,39]. The permeation into the monolayer will decrease when the structure is highly ordered. In this case, when the tmod is 30 min the structure of the
ip t
monolayer allows the maximum permeability of MON and other ions into it.
cr
3.4. Determination of MON accumulation time
us
A study for the determination of the tacc of MON on Au-CA was performed. Cyclic voltammograms were recorded at different tacc in a range from 5 to 160 min to establish the
an
time at which the maximum surface concentration of MON occurs. This study was carried out for MON concentrations in the range between 1 x 10-9 and 1 x 10-7 mol L-1. Plots of Ip,a
M
vs tacc at different concentrations of MON (Fig. 4) showed that the Ip,a increased with the tacc. Then, at about 80 min, a saturation of surface with MON was reached and constant
d
values of Ip,a were obtained. The same result was observed for peak potentials, where they
Ac ce pt e
become less anodic with increasing of tacc and reached a constant value at about 80 min (data not shown). This behaviour is expected from such kind of systems, wherein the electroactive species is accumulated on the electrode surface [40]. In this case, the accumulation is enhanced by both the electrostatic attraction between anionic MON and Au-CA positively charged and the probable formation of a Schiff base. Thus, all measurements were performed at a tacc = 80 min. On the other hand, it was found a linear relationship between anodic and cathodic
charges obtained from oxidation and reduction processes, respectively, for different tacc (Results no shown). Results of linear regression were: slope = 0.57 ± 0.002, intercept = (9.90 ± 0.42) x 10-7 and a linear correlation coefficient (r) of 0.9954.
14
Page 14 of 34
3.5. Calibration curve A calibration curve was constructed from CV results obtained from solutions
ip t
prepared from the commercial reagent. MON standard solutions were prepared in the concentration range from 1 x 10-9 to 1 x 10-7 mol L-1 in CBS, under optimized conditions
cr
previously described. Cyclic voltammograms recorded to construct the calibration curve are
us
shown in Fig. S2 of Supplementary Material. A linear relationship between the anodic peak current and the MON concentration was found. The points used to construct the calibration
an
curve were obtained by triplicate (six points were taken into account, r = 0.9863). The calibration curve is shown in Fig. S3 of Supplementary Material. It is expressed by a least-
M
square procedure as:
(2)
Ac ce pt e
d
I p ,a ( A ) = ( 7.59 ± 0.14 ) x10−7 ( A ) + (4.35 ± 0.28) (A mol L−1 ) c*MON (mol L−1 )
The LOD was 8.3 x 10-10 mol L-1 (0.1 ppb) for a signal-to-noise ratio of 3:1 [41].
The repeatability and reproducibility of the electrochemical sensor was checked
using standard solutions of MON of 2.49 x 10-9 and 2.49 x 10-8 mol L-1. The repeatability was tested by performing three consecutive measurements on the same solution. Values of RSD% were 3.1 % and 3.6 % for 0.3 and 3 ppb, respectively. On the other hand, the reproducibility was obtained from measurement accomplished in three consecutive days. Values obtained were 3.1 % and 3.0 % for 2.49 x 10-9 and 2.49 x 10-8 mol L-1, respectively. These results show that the electrochemical sensor to determine MON has a very good performance.
15
Page 15 of 34
3.6. Analytical determination of MON in maize samples 3.6.1. Determination of MON using the Au-CA electrochemical sensor
ip t
From the extracts obtained as described in Section 2.4, solutions were prepared with both, contaminated and uncontaminated extracts in pH 4 CBS at 10 % v/v, and the cyclic
cr
voltammograms were recorded using the Au-CA sensor (Fig. 5). The solution with the
us
contaminated extract showed an oxidation peak at about 0.45 V vs Ag/AgCl, which can be assigned to the MON oxidation present in the sample. This oxidation peak presented a
an
small potential shift of 0.07 V compared to the peak potential obtained using the commercial reagent. This shift may be caused by matrix effects itself, but we have not
M
found a significant effect in the analytical determination of MON. Regarding to the response obtained with the sensor in the solution with the negative control extract, a wave
d
at about 0.63 V vs Ag/AgCl was observed. This wave is located at higher potentials than
Ac ce pt e
those corresponding to the oxidation peak of MON, and might be assigned to the oxidation of interfering substances present in the matrix. Even when the extract was obtained from an extraction in solid phase using an ESP-SAX column, some components of maize sample could pass through the column and be oxidized in the range of potential used. To minimize the matrix effect, the standard addition method with the Au-CA sensor was used to determine MON in maize samples. A solution with 0.5 % of the contaminated extract in CBS was prepared. This solution was spiked with different amounts of standard solution of MON in order to achieve desired concentrations. The curve obtained for spiked sample was linear in the range from 4.15 x 10-7 to 1.24 x 10-6 mol L-1, with r = 0.9942 and can be represented by at least-square procedure as:
16
Page 16 of 34
(
)
(
I p,a ( A ) = (1.34 ± 0.04 ) x10−6 ( A ) + (1.20 ± 0.05 ) A L mol−1 c*MON mol L−1
)
(3)
10-4 mol L-1.
cr
3.6.2. Determination of MON by HPLC with UV-Vis detection
ip t
Thus, the concentration of MON obtained for the maize sample was (2.23 ± 0.06) x
us
To compare the result obtained from the determination of MON in maize samples using the Au-CA electrochemical sensor, a HPLC/UV-Vis methodology proposed by
an
Parich et al. [12] was performed. The MON concentration determined by HPLC was considered as the reference value.
M
A calibration curve (peak area vs c*MON) was constructed using standard solutions of MON dissolved in the same mobile phase. A linear relationship was found in the
d
concentration range from 4.16 x 10-7 to 1.00 x 10-5 mol L-1 with the following parameters:
Ac ce pt e
intercept = (0.03 ± 0.06), slope = (1.01 ± 0.02) x 106 mol-1 L and r = 0.9991. The LOD was 8.3 x 10-9 mol L-1 (1 ppb) for a 3:1 signal-to-noise ratio and the limit of quantification (LOQ) was 2.49 x 10-8 mol L-1 (3 ppb) for a 10:1 signal-to-noise ratio. On the other hand, a concentration of MON in the contaminated maize sample of (1.92 ± 0.03) x 10-4 mol L-1 was determined by HPLC/UV-Vis. Thus, the concentration value of MON in the contaminated maize sample obtained by both methodologies agrees satisfactorily, indicating the very good performance of the Au-CA electrochemical sensor.
17
Page 17 of 34
Conclusions This is the first report which describes the development of a very sensitive electrochemical sensor for the MON determination in maize samples based on the
ip t
electrochemical oxidation of MON adsorbed at cysteamine self assembled monolayers on gold electrodes. A very good limit of detection of 8.3 x 10-10 mol L-1 (0.1 ppb) and RDS%
cr
less than 4 % in all concentration range were found. The limit of detection obtained was
us
about 400-fold lower than that found using the HPLC method proposed by Parich et al. [12] and others reported in the literature, which shows the good analytical characteristics of the
an
sensor. Also, the measurement of the concentration of MON in maize samples using the sensor and the HPLC methodology were in very good agreement, showing the good
M
performance of the sensor. This Au-CA electrochemical sensor can provide a promising
Ac ce pt e
low cost equipment.
d
prospect for the determination of MON in other real samples using a simple method and a
Acknowledgment
The authors thank Dra. Sofia Chulze for providing real samples of maize and the
HPLC equipment. Financial support from the Consejo Nacional de Investigaciones Científicas y Técnicas (CONICET), Agencia Nacional de Promoción Científica y Tecnológica (FONCYT), Ministerio de Ciencia y Tecnología de la Provincia de Córdoba (MinCyT) and Secretaría de Ciencia y Técnica from the Universidad Nacional de Río Cuarto are gratefully acknowledged. P. C. Díaz Toro and M. V. Fumero thank CONICET for doctoral fellowships. We thank to Reviewers for their valuable suggestions.
18
Page 18 of 34
References [1] J. M. Soriano del Castillo, Micotoxinas en Alimentos, Díaz de Santos, España, 2007. [2] M. Jestoi, Emerging Fusarium – mycotoxins fusaproliferin, beauvericin, enniatins, and
ip t
moniliformin: a review, Crit. Rev. Food Sci. Nutr. 48 (2008) 21-49. [3] V. Scarpino, M. Blandino, M. Negre, A. Reyneri, F. Vanara, Moniliformin analysis in
cr
maize samples from North-West Italy using multifunctional clean-up columns and the
us
LC-MS/MS detection method, Food Addit. Contam. Part A 30 (2013) 876-884. [4] International Agency for Research on Cancer (IARC), 82, 2006.
an
[5] J. A. Engelhard, W. W. Carlton, J. F. Tuite, Toxicity of Fusarium moniliforme subglutinans for chicks, ducklings, and turkey poults, Avian. Dis. 33 (1989) 357-360.
M
[6] D. R. Ledoux, A. J. Bermudez, G. E. Rottinghaus, J. Broomhead, G. A. Bennett, Effects of feeding Fusarium fujikuroi culture material, containing known levels of
d
moniliformin, in young broiler chicks, Poult. Sci. 74 (1995) 297-305.
Ac ce pt e
[7] L. Z. Zhu, Y. M. Xia, G. Q. Yang, Blood glutathione peroxidase activities of populations in Keshan disease affected and non-affected areas, Acta. Nutr. Sin. 4 (1982) 229-233.
[8] P. G. Thiel, A molecular mechanism for the toxic action of moniliformin produced by Fusarium moniliforme, Biochem. Pharmacol. 27 (1978) 483-486. [9] L. T. Burka, J. Doran, B. J. Wilson, Enzyme inhibition and the toxic action of moniliformin and other vinylogous α-ketoacids, Biochem. Pharmacol. 31 (1982) 79-84. [10] M. C. Pirrung, S. K. Nauhaus, B. Singh, Cofactor-directed, time-dependent inhibition of thiamine enzymes by the fungal toxin moniliformin, J. Org. Chem. 61 (1996) 25922593.
19
Page 19 of 34
[11] N.W. Turner, S. Subrahmanyam, S.A. Piletsky, Analytical methods for determination of mycotoxins: A review, Anal Chim Acta 632(2009) 168-180. [12] A. Parich, L.S. Boeira, S.P. Castro, R. Krska, Determination of moniliformin using
ip t
SAX column clean-up and HPLC/DAD-detection, Mycotox. Res. 19(2003) 203-206. [13] C. Munimbazi, L. Bullerman, Chromatographic method for the determination of the
cr
mycotoxin moniliformin in corn, in: M. Trucksess, A. Pohland (Eds.), Mycotoxin
us
Protocols, Humana Press, 2001, pp. 131-145.
[14] G. Filek, W. Lindner, Determination of the mycotoxin moniliformin in cereals by
an
high-performance liquid chromatography and fluorescence detection, J Chromatogr A 732 (1996) 291-298.
Mycotox. Res. 14 (1998) 35-40.
M
[15] F. Schütt, H. Nirenberg, G. Demi, Moniliformin production in the genus Fusarium,
d
[16] M. Jestoi, M. Rokka, A. Rizzo, K. Peltonen, Moniliformin in Finnish grains: Analysis
Ac ce pt e
with LC-MS/MS, Aspect Appl. Biol. 68 (2003) 94-99. [17] C. M. Maragos, Detection of moniliformin in maize using capillary zone electrophoresis, Food Addit. Contam. 21 (2004) 803-810. [18] J. Gilbert, J. R. Startin, I. Parker, M. J. Shepherd, J. C. Mitchell, M. J. Perkins, Derivatization of the Fusarium mycotoxin moniliformin for gas chromatography-mass spectrometry analysis, J. Chromatogr. A 369 (1986) 408-414. [19] P. C. Díaz Toro, F. J. Arévalo, M. A. Zon, H. Fernández, Studies of the electrochemical behavior of moniliformin mycotoxin and its sensitive determination at pretreated glassy carbon electrodes in a non-aqueous medium, J Electroanal. Chem. 738 (2015) 40-46.
20
Page 20 of 34
[20] J. C. Love, L. A. Estroff, J. K. Kriebel, R. G. Nuzzo, G. M. Whitesides, Selfassembled monolayers of thiolates on metals as a form of nanotechnology, Chem. Rev. 105 (2005) 1103-1170.
ip t
[21] H. O. Finklea, Electrochemistry of organized monolayers of thiols and related
Chemistry, Marcel Dekker, New York, 1996, pp. 109-335.
cr
molecules on electrodes, in: A.J. Bard, I. Rubinstein (Eds.), Electroanalytical
Electrochem. Commun. 9 (2007) 2827-2832.
us
[22] M. Sheffer, V. Vivier, D. Mandler, Self-assembled monolayers on Au microelectrodes,
an
[23] H. O. Finklea, S. Avery, M. Lynch, T. Furtsch, Blocking oriented monolayers of alkylmercaptans on gold electrodes, Langmuir 3 (1987) 409-413.
M
[24] P. R. Perrotta, N. R. Vettorazzi, F. J. Arévalo, A. M. Granero, S. N. Chulze, M. A. Zon, H. Fernández, Electrochemical studies of ochratoxin A mycotoxin at gold
d
electrodes modified with cysteamine self-assembled monolayers. Its ultra sensitive
Ac ce pt e
quantification in red wine samples, Electroanalysis 23 (2011) 1585-1592. [25] M. Wanunu, A. Vaskevich, S. R. Cohen, H. Cohen, R. Arad-Yellin, A. Shanzer, I. Rubinstein, Branched coordination multilayers on gold, J. Am. Chem. Soc. 127 (2005) 17877-1787.
[26] S. V. Pereira, F. A. Bertolino, G. A. Messina, J. Raba, Microfluidic immunosensor with gold nanoparticle platform for the determination of immunoglobulin G antiEchinococcus granulosus antibodies, Anal. Biochem. 409 (2011) 98-104. [27] R. K. Shervedani, M. Bagherzadeh, S. A. Mozaffari, Determination of dopamine in the presence of high concentration of ascorbic acid by using gold cysteamine selfassembled monolayers as a nanosensor, Sens. Actuators B 115 (2006) 614-621.
21
Page 21 of 34
[28] M. V. Fumero, M. M. Reynoso, S. Chulze, Fusarium temperatum and Fusarium subglutinans isolated from maize in Argentina, Int. J. Food Microbiol. 199 (2015) 8692.
ip t
[29] D. T. Sawyer, J. L. Roberts, Experimental Electrochemistry for Chemists, 2nd Ed., New York: Wiley, 1995.
cr
[30] J. O. M. Bockris, B. E. Conway, R. E. White, Modern Aspects of Electrochemistry:
us
Springer, 1992.
[31] E. F. Douglass Jr, P. F. Driscoll, D. Liu, N. A. Burnham, C. R. Lambert, W. G. Mc
an
Gimpsey, Effect of electrode roughness on the capacitive behavior of self-assembled monolayers, Anal. Chem. 80 (2008) 7670-7677.
M
[32] J. C. Hoogvliet, M. Dijksma, B. Kamp, W. P. van Bennekom, Electrochemical Pretreatment of polycrystalline gold electrodes to produce a reproducible surface
Ac ce pt e
(2000) 2016-2021.
d
roughness for self-assembly: A study in phosphate buffer pH 7.4, Anal. Chem. 72
[33] B. B. Damaskin, O. A. Petrii, V. V. Batrakov, Adsorption of organic compounds on electrodes, Plenum Press, New York, USA, 1971.
[34] Z. Q. Feng, S. Imabayashi, T. Kakiuchi, K. Niki, Electroreflectance spectroscopic study of the electron transfer rate of cytochrome c electrostatically immobilized on the ω-carboxylalkanethiol monolayer modified gold electrode, J. Electroanal. Chem. 394 (1995) 149-154.
[35] C. P. Smith, H. S. White, Voltammetry of molecular films containing acid/base groups, Langmuir, 9(1993) 1-3. [36] T. W. Solomons, Fundamentals of Organic Chemistry, J. Wiley & Sons, New York, USA, 1982. 22
Page 22 of 34
[37] M. Wirde, U. Gelius, Self-assembled monolayers of cystamine and cysteamine on gold studied by XPS and voltammetry, Langmuir 15 (1999) 6370-6378. [38] M. J. Esplandiú, H. Hagenström, D. M. Kolb, Functionalized self-assembled
ip t
alkanethiol monolayers on Au(111) electrodes: 1. Surface structure and electrochemistry, Langmuir 17 (2001) 828-838.
cr
[39] J. J. Calvente, G. López-Pérez, P. Ramírez, H. Fernández, M. A. Zon, W. H. Mulder,
us
R. Andreu, Experimental study of the interplay between long-range electron transfer and redox probe permeation at self-assembled monolayers: Evidence for potential-
an
induced ion gating, J. Am. Chem. Soc. 127 (2005) 6476-6486.
[40] J. C. Love, D. B. Wolfe, R. Haasch, M. L. Chabinyc, K. E. Paul, G. M. Whitesides, R.
M
G. Nuzzo, Formation and structure of self-assembled monolayers of alkanethiolates on palladium, J. Am. Chem. Soc. 125 (2003) 2597-2609.
d
[41] D. MacDougall, W. B. Crummett, Guidelines for data acquisition and data quality
Ac ce pt e
evaluation in environmental chemistry, Anal. Chem. 52 (1980) 2242-2249.
23
Page 23 of 34
Figure Captions Figure 1. Chemical structure of moniliformin. X = Na, K or H.
ip t
Figure 2. Cyclic voltammograms recorded in pH 4 CBS at Au-CA (A) and Au-2-DAET (B) modified gold electrodes in blank (1) and in MON (2) solutions. c*MON = 2.4 x 10-4 mol
us
cr
L-1. cthiol = 2 x 10-3 mol L-1. tmod= 60 min. tacc = 80 min. v = 0.100 V s-1.
Figure 3. Variation of the anodic peak current with: (A) the concentration of cysteamine
an
for a modification time of 60 min, and (B) the modification time at a cysteamine concentration of 2.5 x 10-3 mol L-1 (B). In both experiments the concentration of MON was
M
2.4 x 10-4 mol L-1. The dash line was plotted to show the trend.
d
Figure 4. Variation of the anodic peak current with the accumulation time (tacc) at three
Ac ce pt e
concentration levels of MON. c*MON = 1 x 10-9 (1), 5 x 10-8 (2), 1 x 10-7 mol L-1 (3). The dash line was plotted to show the trend.
Figure 5. Cyclic voltammograms recorded with the Au-CA sensor in a solution with the negative control (1) and the contaminated extract (2). Both solutions were prepared in 10 % v/v pH 4 CBS. v = 0.100 V s-1.
Figure S1. Cyclic voltammograms of the blank (1) and MON in pH 4 CBS (2) solutions. c*MON = 2.4 x 10-4 mol L-1. v = 0.100 V s-1. Working electrode: bare gold electrode.
24
Page 24 of 34
Figure S2. Cyclic voltammograms recorded to construct the calibration curve using the Au-CA sensor in solutions of MON + pH 4 CBS. MON concentrations were: 1) 1.00 x 10-9;
Ac ce pt e
d
M
an
us
cr
Figure S3. The calibration curve obtained with data from Fig. S2.
ip t
2) 2.50 x 10-9; 3) 3.33 x 10-8; 4) 5.00 x 10-8 and 5) 1.00 x 10-7 mol L-1. v = 0.100 V s-1.
25
Page 25 of 34
Vitae Paulo César Díaz Toro obtained his Ph. D. in Chemistry (2015) from Río Cuarto National University (Río Cuarto, Argentina). He obtained a fellowship from Argentine Research
ip t
Council (CONICET) and was an active member of the Electroanalytical Group at the Chemistry Department, Faculty of Exact, Physico-Chemical and Natural Sciences (Río
cr
Cuarto National University). His research interest focuses on the development of
us
electroanalytical techniques for the determination of mycotoxins as well as the design and
an
characterization of electrochemical sensors.
Fernando J. Arévalo obtained his Ph. D. in Chemistry (2009) from Río Cuarto National
M
University (Río Cuarto, Argentina). He is a Researcher at Argentine Research Council (CONICET). At present, he also is Assistant Professor at the Chemistry Department,
d
Faculty of Exact, Physico-Chemical and Natural Sciences (Río Cuarto National
Ac ce pt e
University). Dr. Arévalo is an active member of the Electroanalytical Group at the Chemistry Department, and his research interest focuses on the development and characterization of electrochemical (bio) sensors based on the use of nano-structured materials.
María Verónica Fumero obtained her graduate in Microbiology (2012) from Río Cuarto National University (Río Cuarto, Argentina). She is actually doing a Ph. D. in Biological Sciences in the Microbiology and Immunology Department, Faculty of Exact, PhysicoChemical and Natural Sciences (Río Cuarto National University). At present, she has a doctoral fellowship from Argentine Research Council (CONICET) at the same Department. She is an active member of the Mycology and Mycotoxicology Group at the Microbiology 26
Page 26 of 34
and Immunology Department, and her research interest focuses on the study of species of Fusarium pathogens of maize and producers of diverse mycotoxins.
ip t
María A. Zon obtained her Ph. D. in Chemistry (1985) from Río Cuarto National University (Río Cuarto, Argentina). She did the postdoctoral training at Cordoba University
cr
(Córdoba, España) between 1990 and 1992. She is Full Professor at the Río Cuarto
us
National University and Principal Researcher at Argentine Research Council (CONICET). She has been the secretary of the Analytical Chemist Argentine Association (2007-2009).
an
Her research interest focuses on the development of electrochemical (bio) sensors by using nano-materials for the determination of different substrates such as mycotoxins, natural
M
antioxidants and hormones. She has over 65 peer-reviewed papers and eight book chapters. She is co-author of a book. She has been co-editor of an electroanalytical book. Prof. Zon is
Ac ce pt e
d
an AAQA, AAIFQ and SIBAE fellow.
Héctor Fernández obtained his Ph. D. in Chemistry (1978) from Río Cuarto National University (UNRC) (Río Cuarto, Argentina). He did the postdoctoral training (1980-1982) at the University of New York at Buffalo, Buffalo (USA). Currently, he is Full Professor at UNRC and Principal Researcher at Argentine Research Council (CONICET). He was Dean of the Faculty of Exact, Physico-Chemical and Natural Sciences (UNRC, 1992-1999) and Head of the Department of Chemistry at the Faculty of Exact, Physico-Chemical and Natural Sciences (2001-2004). He was President of the Argentinean Society of Analytical Chemists (2007-2009). His research interest focuses on several subjects, such as electrochemistry of mycotoxins, hormones and synthetic and natural antioxidants, studies on ultramicroelectrodes and electrodes modified by self-assembled monolayers of thiols, 27
Page 27 of 34
carbon nanotubes, antibodies, etc and their use for electroanalytical applications. Development of electroanalytical techniques for the determination of antioxidants, mycotoxins and hormones in real matrixes (plants, cereal, foods, sera of animal origin, etc,
ip t
respectively). Design and characterization of chemical sensors, electrochemical (bio)sensors and immunoelectrodes based on nanostructured materials. He has over eighty
cr
peer-reviewed papers, eight book chapters, co-author of a book and has been the editor of a
us
book. Prof. Fernández belongs to the Editorial Board of J Biosensors and Bioelectronics and Polish Journal of Environmental Studies. He is an AAQA, AAIFQ, SIBAE and ISE
An electrochemical sensor was developed to determine moniliformin in maize
•
Ac ce pt e
samples.
d
•
M
an
fellow.
A self-assembled monolayer modified gold electrode was used as working electrode.
• • •
Cyclic voltammetry was the electrochemical technique used. The calibration curve was linear in the concentration range from 0.12 to 12 ppb. Results of the electroanalytical method were in good agreement with those of HPLC.
28
Page 28 of 34
ed pt ce Ac
Page 29 of 34
a M ed pt ce Ac
Page 30 of 34
M a d te ep Ac c
Page 31 of 34
us an M ed pt ce Ac
Page 32 of 34
Page 33 of 34
ed
ce pt
Ac M an
c
us
ed pt ce Ac Page 34 of 34