A C8H8O3) is one of the world’s flavour

A new chemically modified electrode
based on titanium dioxide nanoparticles (TiO2-NPs) has been developed.
Aluminium was incorporated into the TiO2-NPs to prepare aluminium
doped TiO2 nanoparticles (Al-TiO2-NPs). Aluminium doped
TiO2 nanoparticles-modified screen printed carbon electrode (Al-TiO2-NPs/SPCE)
was employed as easy, efficient and rapid sensor for electrochemical detection
of vanillin in various types of food samples. Al-TiO2-NPs were
characterized by energy-dispersive X-ray (EDX), transmission electron
microscopy (TEM), and X-ray diffraction (XRD) and analyses showing that the
average particle sizes varied for the Al-NPs (7.63 nm) and Al-TiO2-NPs
(7.47 nm) with spherical crystal. Cyclic voltammetry (CV) and linear sweep
voltammetry (LSV) and were used to optimize the analytical procedure. A
detection limit of vanillin was 0.02 µM, and the relative standard deviation (RSD)
was 3.50 %, obtained for a 5.0 µM concentration of vanillin. The
electrochemical behaviour of several compounds, such as vanillic acid, vanillic
alcohol, p-hydroxybenzaldehyde and p-hydroxybenzoic, etc., generally present in
natural vanilla samples, were also studied, to check the interferences with
respect to vanillin voltammetric signal. The applicability was demonstrated by
analysing food samples. The obtained results were compared with those provided
by a previous method based on liquid chromatography for determination of

(4-hydroxy-3-methoxybenzaldehyde, C8H8O3) is
one of the world’s flavour extracts obtained primarily from Vanillia, a specie
of tropical climbing. Although more than 12,000 tons of vanillin are produced every
year, the natural vanillin from Vanilla is less than 1%; the remainder is
synthesized much more cheaply via biochemical and/or chemical processes 1. For the time being, many
analytical methods have been proposed for vanillin determination in various
types of food samples or vanilla extracts, including fluorescence 2, capillary electrophoresis (CE) 3, liquid chromatography 4, and GC-MS 5. These have high cost and involve time-consuming sample
pretreatment processes. Because vanillin is an electro-active compound and it
is possible to quantify the vanillin amount in vanilla and in the final
products by electrochemical detection (ECD) through the study of its oxidation.
ECD is important method for quantitative determination of vanillin due to their
easy to use, fast response, high sensitivity, and cheap instrumentation 6–11. Various electrochemical methods, such as amperometry, differential
pulse voltammetry (DPV) or square-wave voltammetry (SWV) for the determination
and detection of vanillin in various food samples have been reported in
literature 12,13.

In recent years, many
reports on screen-printed carbon electrodes (SPCEs) technology have been used
to develop various electrochemical sensors that detect molecules in various
sectors, such as biomedical environmental and agri-food 14,15. SPCE is miniaturized and planar shape, thus can be used as drop on
sensor, using few microliters of the sample. Moreover, it is low-cost and can
be used as a disposable electrode with large-scale production capability. The
oxidation of vanillin at SPCE and others most unmodi?ed electrode surfaces
represents a serious problem which arises from the high over-potential together
with poor reproducibility resulted from a fouling effect, which cause rather
poor selectivity and sensitivity 16,17. In order to avoid these disadvantages, the modification of
electrochemical working electrode is an excellent alternative. In this sense,
many researchers have attempted to diminish the over-potentials by using various
modi?ed electrodes such as, graphene 6,18, silver nanoparticles (Ag-NPs) 8, gold nanoparticles (Au-NPs) 9,10, and multi-walled carbon nanotubes (MWCNTs)
19. However, the performance of some
electrodes was still not enough good. Thus it is
important to develop new types of electrode devices by the synthesis and
preparation of new materials (modi?ers). In this regard, as electrode surface
modi?ers, titanium dioxide nanoparticles (TiO2-NPs) captured great interest
of researchers and a considerable amount of research was executed in the
previous decades 12,13,19. The augment in the ef?ciency of TiO2-NPs such as to ensnare
the charge carrier is achieved by doping with transition metals and transition
metal oxides such as, copper (Cu) 20, cadmium oxide nanoparticles (CdO-NPs) 21, silver nanoparticles (Ag-NPs) 22, gold nanoparticles (Au-NPs) 23 and ruthenium (Ru) 24.

Herein we report
the synthesis of aluminium doped TiO2 nanoparticles (Al-TiO2-NPs),
which were characterized by TEM and XRD. The sensitivity of the developed
electrode is compared with SPCE, TiO2-NPs/SPCE, and Al-TiO2-NPs/SPCE
for vanillin detection. The results show that a composite film of Al-TiO2-NPs/SPCE
is more sensitive compared to SPCE and TiO2-NPs/SPCE. Further, we
investigated the electrochemical behaviour of vanillin at modified SPCE with
Al-TiO2 (Al-TiO2-NPs/SPCE). The performance of the
modified sensor is demonstrated by the determination of vanillin in food
samples obtaining good selectivity, stability and high sensitivity.

2.All the starting
materials were purchased with very highly purity. Aluminium acetylacetonate
(Al(acac)3, 99%), lithium aluminium hydride (LiAlH4,
95%), mesitylene (97%), titanium (IV) oxide (anatase, powder, 235 mesh),
4-hydroxybenzaldehyde (98%) and 4-hydroxy-3-methoxybenzyl alcohol (98%) were
purchased from Sigma-Aldrich (St. Louis, MO, USA). Nafion 117 solution (5% in a
mixture of lower aliphatic alcohols and water), vanillin, vanillic acid,
4-hydroxybenzoic acid were purchased from Fluka Chemie (Buchs, UK). Ethanol and
phosphoric acid were purchased from Panreac Quimica S.L.U. (Barcelona, Spain).
Vanillin solution were prepared in ethanol and stored in the dark until use.
Working standard solutions were made by appropriate dilution of the vanillin
stock standard solution with phosphate buffer at pH 6.3. Water was puri?ed with
a Milli-Q system (Millipore). Ethanol and phosphoric acid were obtained from
Panreac Quimica S.L.U. (Barcelona, Spain).

Vanillin extract
samples (sample A and sample B) were purchased from different local markets
(Ciudad Real, Spain). These extracts were filtered through a sintered filter,
and diluted directly in phosphate solution.

detection was carried out on a CH Instruments Model 800D Series (Austin, Texas,
USA) all measurements were carried out using a screen printed carbon electrodes
(SPCEs) system (Dropsens DRP-C110) housed in the home made electrochemical flow
cell. Transmission electron microscopy (TEM) micrographs were measured on a
Jeol JEM 2011 operating at 200 kV and equipped with an Orius Digital Camera (2
× 2 MPi). The samples were prepared by deposition of a drop of the synthesized
material suspension onto a lacey carbon/format-coated copper grid. The digital
analysis of the HRTEM micrographs was done using Digital Micrograph TM 1.80.70 for
GMS 1.8.0 Gatan. XRD patterns were measured on Philips model X´Pert MPD
diffractometer using a CuKa source (? =1.5418 Å), programmable divergence slit,
graphite mono-chromator and proportional sealed xenon gas detector. The samples
were made with a voltage of 40 KV, intensity 40 mA and an angular range of 20
to 70 degrees (2?), a step of 0.02
degrees 2 Theta and a time per step of 1.50 sec.

Agilent 1200
liquid chromatography system was used as a chromatographic system. It was
equipped with a LC pump, a vacuum degasser, a micro well-plate auto sampler (5
?L injection loop), a thermostatted column compartment and a Diode-Array
Detector (DAD). The data was processed using the Agilent ChemStation software.
Reversed-phase C18 analytical column Luna 5µm PFP (2) 100A (150
x 4.6 mm) was used for the separation of the analytes present in vanillin
extract samples. Elution was performed under isocratic conditions, by using a
mixture of acetonitrile/phosphate 20/80 (v/v) as a mobile phase, at a flow-rate
of 1.0 mL min?1. The injection volume was 40 µL. Detection wavelength
was performed at 265 nm.

 At first, Al-NPs
were prepared according the previously described procedures 25, incorporating some modifications. This modified synthesis involves
the addition of aluminium acetylacetonate Al(acac)3 (10 mmol) to
mesitylene which was already placed in a three-neck round bottom flask with
magnetic bar. Then lithium aluminium hydride LiAlH4 (30 mmol) was
added to the mixture. The reaction was purged with N2 during reflux
with stirring for 72 hours at 165 °C. The
reaction mixture was cooled down to 25 °C. A gray-colored precipitate was crushed and dried under low
pressure for 5 hours. The crude product was washed with 25 ml of cold methanol
at least three times, to avoid any exothermic reaction between solvent and
Al-NPs. The unreacted starting materials were washed with methanol three times.
The final product was filtered and dried at 25 °C under low pressure. The preparation of Al-TiO2-NPs
started by mixing of TiO2 (0.5 g) previously digested in nitric acid
(0.1 M, 25 mL) during 3 hours and Al-NPs (0.5 g) previously prepared. The final
product was filtered and dried at 25 °C under low
pressure, obtaining an Al-TiO2-NPs as light-gray powder.

were dispersed in water (0.5% Nafion, v:v) by ultrasound sonicator, obtaining
individual concentrations of 1 mg mL-1. Films formed from
nafion-solubilized nanoparticles are more uniform than those casted by organic
solvents 26. TiO2-NPs and Al-TiO2-NPs were used. DRP-110
SPCEs (Dropsens, Spain), with carbon as a working electrode, were used to
prepare the modi?ed-SPCE. Each modi?ed SPCE was prepared by casting 2 µL of the
dispersed NPs onto the surface of the electrode. After drying under infrared
light (IR) for 15 min, rinsing with pure water and ready for using. Energy-dispersive
X-ray (EDX) elemental mapping and transmission electron microscopy (TEM)
micrograph for the aluminium and aluminium doped titanium nanoparticles (Al-TiO2-NPs)
are shown in Figure 1. The Figure shows that different materials have different
surface morphologies. Figure 1 (A and C), shows the TEM and EDX micrographs of
the aluminium particles without doping (this image was obtained before
aluminium was mixed with TiO2-NPs.), and a higher magnification is
also presented. A large number of precipitates were distributed homogeneously through
the aluminium grains in the peak aged conditions. It is clear that the increase
in hardness is brought about by the distribution of the precipitates for
aluminium. Figure 1 (B and D), shows the TEM and EDX micrographs of the fully
mixed precursor. It shows two large particles in the lower region, which are
Al-NPs and TiO2-NPs, were distributed homogeneously, indicating that
a large amount of Al-NPs has been doped in the TiO2-NPs. Figure 2,
shows XRD patterns measured for Al-NPs (Figure 2 (A)) and Al-TiO2-NPs
(Figure 2 (B)). The XRD patterns show that the nanoparticles involved aluminium
as a major component. The XRD patterns and all the positions of the peaks are found
to be in good agreement with the face-centered cubic (fcc) form of aluminium 25, and anatase TiO2 27, as shown in Figure 2. A closer look at the figure shows that the
Al-TiO2-NPs involved aluminium as a major component mixed with TiO2-NPs
as minor components, as shown in Figure 2 (b) 28. Calculations based on the Scherrer equation (D=K?/?cos?) 29, show that the average particle sizes varied for the Al-NPs (7.63
nm) and Al-TiO2-NPs (7.47 nm). The XRD results confirm the TEM
micrograph results discussed above.

The modified screen
printed carbon electrodes (SPCEs) by aluminium doped TiO2
nanoparticles (Al-TiO2-NPs/SPCE) were ?rst characterized by cyclic
voltammetry (CV) to test their behaviour for the oxidation of vanillin, as
shown in Figure 3 (C). In order to find the role of Al-NPs, the cyclic
voltammograms of (A) SPCE, (B) TiO2-NPs/SPCE, and (C) Al-TiO2-NPs/SPCE
in the presence of 250 µM vanillin were recorded, as shown in Figure 3. The
electrochemical behaviours of the TiO2-NPs/SPCE and Al-TiO2-NPs/SPCE
were investigated by using the CV technique, as shown in Figure 3 (B and C).
All cyclic voltammograms were obtained in the presence of 0.1 M H3PO4
electrolyte (pH=6.3) with 250 µM vanillin at 50 mV s-1 scan rate.
The cyclic voltammograms observed for the SPCE electrode (with the absence of
TiO2-NPs and Al-TiO2-NPs, Figure 3 (A)). In contrast, the
cyclic voltammograms observed for the TiO2-NPs/SPCE and Al-TiO2-NPs/SPCE
modified electrode exhibited only an oxidation peak in the presence of
vanillin, within the potential window in the range from 0.00 to 1.20 V. It
suggests that the oxidation reaction of vanillin on Al-TiO2-NPs/SPCE
is totally irreversible. The oxidation peak of vanillin is observed at the SPCE
(Figure 3 (A)) with a low and broad peak current. The oxidation current of
vanillin (0.58 V) on Al-TiO2-NPs/SPCE (Figure 3 (C)) was higher than
that on the SPCE. Compared with the SPCE, TiO2-NPs/SPCE and Al-TiO2-NPs/SPCE,
a significant enhancement in the anodic current (0.58 V) was achieved at the
Al-TiO2-NPs/SPCE (Figure 3 (C)), indicating that the high
conductivity and high surface area of the TiO2-NPs/SPCE improve the
catalytic activity and increase the effective electrode area toward the
vanillin oxidation obviously indicating that the Al-TiO2-NPs/SPCE
can be used to determine vanillin.

electrochemical response of the modified Al-TiO2-NPs/SPCE toward the
determination of vanillin were optimized by analyzing a standard solution (10
µM) of vanillin using linear sweep voltammetry (LSV) technique. The parameters
a?ecting the determination of vanillin, such as electrolyte, pH,
Al-TiO2-NPs amount, scan rates, adsorption time, accumulation
conditions and stability of the electrodes, were investigated. In order to find
the optimal parameters, one parameter was changed and the other parameters were
?xed at their optimal values.

The influence of
different supporting electrolytes were tested, including NH4Ac, HCl,
H3PO4, HNO3, H2SO4 and
NaOH (each 0.1 M). The results indicated that when 0.1 M H3PO4 solution
was used, the oxidation peak current was a higher sensitivity than other. For
further study, 0.1 M H3PO4 solution was chosen as the
supporting electrolyte. On the other hand, the effect of 0.1 M H3PO4
supporting electrolyte pH on the electrochemical response of the modified Al-TiO2-NPs/SPCE
toward the determination of vanillin was studied. The variations of the oxidation
peak potential as well as the peak current with respect to the change of the
electrolyte in the pH range (1.40 – 6.30), with an
increase the pH, manifesting that protons have taken part
in the electrode reaction processes. The relationship between the oxidation
peak potential and pH is also shown in Figure S1. The effect of modification
amount of Al-TiO2-NPs on the SPCE surface for the determination of
vanillin was also studied by using LSV, different modification volumes were
tested: 2.0, 4.0 and 6.0 ?g. The 2.0 ?g gave the best result of oxidation peak
current sensitivity than other counterparts. The effect of scan rate value on
the oxidation peak current of 10 µM of vanillin by using LSV was studied.
Different scan rates were tested 20, 30, 50 and 70 mV s-1. The 50 mV
s-1 gave the best result of oxidation peak current sensitivity than
other counterparts. In this study, 50 mV s-1 scan rate was used. The
accumulation step of 10 µM vanillin after 120 second was performed under 0.0 V,
the peak current increased progressively with accumulation time up to 120
second at a fixed accumulation potential 0.0 V. Thereafter, the peak current
increased much slightly as further increasing the accumulation time, also shown
in Figure S2. This phenomenon could be attributed to the saturated adsorption
of vanillin on the surface of the electrode. Then, the stability of the
modified electrodes was examined under the best conditions. Five electrodes
were made by the same procedure. Moreover, the stability of modified electrode
was also studied. When the modified electrode (Al-TiO2-NPs/SPCE) was
studied with eight segments, the peak current kept 99.0 % of the original peak
is also shown in Figure S3. Using the
previously optimized conditions, the analytical parameters of the system were
examined by linear sweep voltammetry (LSV), using a scan rate of 50 mV s-1.
A linear calibration graph was obtained using standard vanillin solutions in
the range of 0.07–20 µM. The linear sweep voltammograms obtained for different
concentrations of vanillin under the optimum experimental conditions. The
linear range, slope and intercept of the calibration curve are given in Table 1,
along with the regression coefficient (R2) for vanillin. The
precision of the method for aqueous standard solutions (evaluated as the
relative standard deviation (RSD) obtained after analyzing 10 series of 10
replicates) was 3.5 % at the 5 µM concentration of vanillin. The theoretical
limit of detection (LOD), is expressed as the analyte concentration giving a
signal equivalent to the blank signal plus three times its standard deviation
(3?), was 0.02 µM. This method provides clear and good advantages
in terms of sensitivity with respect to other existing alternatives methods 16,17 that involve the use of others vanillin sensors and electrochemical
detection (see Table 2).

The influence of
potential interferences in the analytical signal of vanillin was studied for
some common compounds in the samples to be analyzed, such as vanillic acid,
vanillic alcohol, p-hydroxybenzaldehyde and p-hydroxybenzoic acid. These
compounds were chosen because they are known to be present in natural vanilla. A
compound was considered as interference if it caused an analytical variation of
more than 5% when compared to the analytical signal obtained in the absence of
the interfering compound. All the results are shown in Table 3. No
interferences were observed for the interferent/analyte ratios investigated.

The applicability
of the proposed methodology was evaluated by determining vanillin in two
vanillin extract samples (Sample A and Sample B). These vanillin extract
samples were purchased from local markets and prepared according the section 2.1.
Sample A and Sample B were found to contain vanillin. The results obtained are
shown in Table 4. This was compared with a blind analysis of the vanillin
extract samples using the modified HPLC-DAD method 30 described in the section 2.2, which provided a chromatogram
represented in Figure 4. The results showed, both methods are in close
agreement for supporting the validity of this method. In summary, the development of an
aluminium doped TiO2 nanoparticles have been carried out, and the
obtained hybrid nanoparticles have been tested in SPCE, as novel working
electrode for detection of vanillin in extract vanilla samples with good analytical
performance. The use of the developed working electrode (Al-TiO2-NPs/SPCE)
allows the simple, effective and rapid method for electrochemical detection of
vanillin in extract vanilla samples. The Al-TiO2-NPs/SPCE electrode was
found to increase the sensitivity higher than the commercial electrodes tested,
therefore evaluating the effectiveness of the proposed approach to improve the
sensitivity of the working electrodes in screen printed carbon electrodes.