Applied Surface Science 347 (2015) 40–47

Contents lists available at ScienceDirect

Applied Surface Science

journal homepage: www.elsevier.com/locate/apsusc

Hybrid methyl green/cobalt-polyoxotungstate nanostructured films:

Self-assembly, electrochemical and electrocatalytic properties
Hugo C. Novais, Diana M. Fernandes ∗ , Cristina Freire ∗
REQUIMTE, Departamento de Química e Bioquímica, Faculdade de Ciências, Universidade do Porto, 4169-007 Porto, Portugal

a r t i c l e i n f o a b s t r a c t

Article history: Hybrid multilayer films were prepared by alternately depositing cationic dye methyl green (MG) and
Received 6 December 2014 anionic sandwich-type polyoxometalate K10 [Co4 (H2 O)2 (PW9 O34 )2 ] (Co(PW9 )2 ) via electrostatic layer-by-
Received in revised form 28 March 2015 layer (LbL) self-assembly method. Film build-up was monitored by UV–vis spectroscopy which showed
Accepted 5 April 2015
a regular stepwise growth. X-ray photoelectron spectroscopy data confirmed the successful fabrication
Available online 17 April 2015
of the hybrid films with MG-Co(PW9 )2 composition and scanning electron microscopy images revealed
a completely covered surface with a non-uniform distribution of the molecular species. Electrochem-
Keywords:
ical characterization of films by cyclic voltammetry revealed two tungsten-based reduction processes
Layer-by-layer self-assembly
Sandwich polyoxometalate
in the potential range between −0.9 and −0.5 V due to WVI → WV in Co(PW9 )2 . Studies with the redox
Organic dye probes, [Fe(CN)6 ]3−/4− and [Ru(NH3 )6 ]3+/2+ , revealed that not only the electrostatic attractions or repul-
Electrocatalysis sions have effects on the kinetics of the probe reactions, but also the film thickness. Additionally, the
{MG/Co(PW9 )2 }n multilayer films exhibit efficient W-based electrocatalytic activity towards reduction
of nitrite and iodate.
© 2015 Elsevier B.V. All rights reserved.

1. Introduction method, passive diffusion, electrodeposition and self-assembly

[16–24]. Due to its simplicity and flexibility, the LbL self-assembly
Polyoxometalates (POMs) are a large family of inorganic metal- method is the most widely used method for fabricating mul-
oxygen cluster compounds which have been applied in different tilayer films and is based upon the electrostatic attractions
fields, such as material science, magnetochemistry, catalysis and between oppositely charged moieties [25–27]. It results in highly
electrocatalysis [1–5]. Among them are the POMs with sandwich uniform and stable films that possess a high degree of struc-
structure, [M4 (H2 O)2 (PW9 O34 )2 ]10− , M = transition metal, which tural and functional control and can be applied to substrates
are based on the tri-lacunary derivative of the parent Keggin like carbon, glass, silicon and quartz [17,26,27]. A variety of
structure. The sandwich-type anion consists of two lacunary, B-␣- charged materials have be employed for the construction of LbL
[PW9 O34 ]9− Keggin moieties linked via four substituted transition films, including polycations, such as, poly(allylamine hydrochlo-
metal cations. This specific type of sandwich polyoxometalates ride) (PAH), poly(diallyldimethylammonium chloride) (PDDA) and
have been mainly used as catalysts, in particular, the iron and poly(ethylenimine) (PEI) [28–33], polymers, such as, chitosan
manganese sandwich complexes [6,7]. Some studies on the elec- and polypyrrole [34,35], transition metal complexes [25,36] and
trochemical behavior of [M4 (H2 O)2 (PW9 O34 )2 ]10− for M = Cu, Mn, nanoparticles [37,38].
Fe and Co in non-aqueous media can also be found in the literature In this work, we report the successful fabrication of hybrid
[6,8–10]. More recently, the cobalt sandwich complex has gained multilayer films based on two molecular species: the cationic
particular attention as a water oxidation electrocatalyst [11–15]. dye methyl green and the anionic sandwich-type polyox-
However, the application of POMs as functional materi- ometalate K10 [Co4 (H2 O)2 (PW9 O34 )2 ], using the electrostatic LbL
als depends on their effective surface immobilization. Various self-assembly method. The film growth was monitored by UV–vis
strategies have been employed, including, conducting polymers, spectroscopy, the surface composition was evaluated by X-ray
Langmuir-Blodgett and sol-gel films, the layer-by-layer (LbL) photoelectron spectroscopy and surface morphology was exam-
ined by scanning electron microscopy (SEM). The electrochemical
behaviour of the multilayer films and film permeability towards the
∗ Corresponding authors. Tel.: +351 2204020590; fax: +351 220402695.
redox probes [Fe(CN)6 ]3−/4− and [Ru(NH3 )6 ]3+/2+ was studied by
cyclic voltammetry and finally electrocatalytic properties towards
E-mail addresses: diana.fernandes@fc.up.pt (D.M. Fernandes), acfreire@fc.up.pt
(C. Freire). nitrite and iodate reduction was evaluated.

http://dx.doi.org/10.1016/j.apsusc.2015.04.035
0169-4332/© 2015 Elsevier B.V. All rights reserved.
 H.C. Novais et al. / Applied Surface Science 347 (2015) 40–47 41

2. Experimental slides for XPS and on glassy carbon for cyclic voltammetry studies.
Quartz slides were cleaned by placing them in a H2 SO4 /H2 O2 (3:1)
2.1. Reagents and solutions (v/v) hot bath (∼80◦ C) for 30 min and then in a H2 O/H2 O2 /NH3
(5:1:1) (v/v/v) hot bath (∼80◦ C) for 30 min. The cleaned quartz
Sodium acetate (99.5%, Merck), sodium nitrite (99%, Merck), slides were then rinsed with ultra-pure water and dried under
potassium iodate (99.5%, BDH chemicals), acetic acid (100%, a stream of argon. After the cleaning step, the quartz slide
Merck), methyl green (zinc chloride salt, <0.5% crystal violet, was immersed initially in a solution of PEI (0.12 mol dm−3 ) for
dye content 80%) were used as received. Poly(ethylenimine) 10 min, followed by rising with water and dried under argon.
(MW = 50,000–100,000; 30 wt.% aqueous solution; branched, con- The PEI-coated quartz slides were then alternately immersed into
sisting of tertiary, secondary and primary amino groups in the ratio Co4 (PW9 )2 solution (5.0 × 10−3 or 1.0 × 10−3 mol dm−3 ) for 10 min
of 25/50/25, respectively) was purchased from Polysciences Europe and MG solution (1.0 × 10−3 or 5.0 × 10−4 mol dm−3 ) for 10 min,
GmbH and used without further treatment. also rinsed with water and dried under argon after each dipping
Potassium hexacyanoferrate (III) and hexaammineruthenium step. This procedure was repeated as many times as the number of
(III) solutions (1.0 mmol dm−3 ) were prepared by dissolving bilayers needed. Deposition times of 5 and 20 min were also tested.
the appropriate amount of K3 [Fe(CN)6 ] and [Ru(NH3 )6 ]Cl3 in All solutions were prepared using a pH 4.0 acetate buffer solution
1 mol dm−3 KCl. prepared as described in Section 2.1.
Electrolyte solutions for cyclic voltammetry were prepared For the glass slides the same initial solution treatments were
using ultra-pure water (resistivity 18.2 M cm at 25 ◦ C, Millipore). used, but for just 1–2 min. The GCE was conditioned by a pol-
Acetate buffer solutions with pH 4.0 used to prepare the LbL ishing/cleaning procedure prior to coating: it was polished with
films and for electrochemical studies of these films were prepared aluminium oxide of particle size 0.3 ␮m (Buehler-Masterprep) on a
by mixing appropriate amounts of CH3 COOH 0.1 mol dm−3 and microcloth polishing pad (BAS Bioanalytical Systems Inc.), and then
NaCH3 COO 0.1 mol dm−3 solutions. the electrode was rinsed with ultra-pure water and finally sonica-
ted for 5 min. The multilayer films were build-up on the surface
2.2. Instrumentation of glass slides and GCE electrodes using the same methodology as
described above for the quartz slides: initially they were immersed
Electronic spectra of the multilayer films were recorded on a in a 0.12 mol dm−3 PEI solution for 10 min, then immersed in a
Varian cary 50 Bio spectrophotometer in the range 190–1000 nm, 1.0 × 10−3 mol dm−3 Co4 (PW9 )2 solution for 10 min and finally in
at room temperature, using quartz slides. 5.0 × 10−4 mol dm−3 MG for 10 min.
Scanning electron microscopy (SEM) analysis was performed
at the Centro de Materiais da Universidade do Porto (CEMUP),
Portugal, in high-vacuum conditions, using a high-resolution 3. Results and discussions
(Schottky) environmental scanning electron microscope (ESEM).
Before SEM measurements a thin Au layer was deposited on the 3.1. {MG/Co4 (PW9 )2 }n film build-up
top of the samples.
The X-ray photoelectron spectroscopy (XPS) measurements UV–vis spectroscopy was used to monitor multilayer build-up
were performed at CEMUP (Porto, Portugal) using a Kratos AXIS of PEI/Co4 (PW9 )2 /{MG/Co4 (PW9 )2 }n films. The electronic spectrum
Ultra HSA, with VISION software for data acquisition and CASAXPS of MG in pH 4.0 acetate buffer solution (Figure S1(a)) shows one
software for data analysis. The analysis was carried out with a intense band at max = 632 nm and two other bands at max = 422
monochromatic Al K␣ X-ray source (1486.7 eV), operating at 15 kV and 312 nm. The electronic spectra of Co4 (PW9 )2 (Figure S1(b))
(90 W), in FAT mode (Fixed Analyser Transmission), with a pass shows in pH 4.0 acetate buffer solution the characteristic high
energy of 40 eV for regions ROI and 80 eV for survey. Data acquisi- intense electronic bands in the region 190–300 nm: the electronic
tion was performed with a pressure lower than 1.E−6 Pa, and it was band at max = 252 nm is characteristic of Keggin structure and is
used a charge neutralisation system. To correct possible deviations attributed to a ligand-to-metal-charge transfer transition (LMCT)
caused by electric charge of the samples, the C 1s band at 285.0 eV from the bridging oxygen atoms Ob and Oc to WVI atoms (Ob(c) → W)
was taken as internal standard. The XPS spectra were deconvo- and the band at max ≈ 200 nm is attributed to the LMCT transition
luted with the XPSPEAK 4.1 software, using non-linear least squares from the terminal oxygen atoms to the WVI atoms (Ot → W)[39];
fitting routine after a Shirley-type background subtraction. The sur- this latter electronic band is difficult to observed due to the solvent
face atomic percentages were calculated from the corresponding cut-off.
peak areas and using the sensitivity factors provided by the manu- Fig. 1 depicts the UV–vis spectra of {MG/Co4 (PW9 )2 }n film (with
facturer. n = 1–15) sampled after deposition of an even number of layers
Cyclic voltammetry measurements were carried out using an during its assemblage on a quartz substrate using 0.12 mol dm−3
Autolab PGSTAT 30 potentiostat/galvanostat (EcoChimie B.V.) con- of PEI, 1.0 × 10−3 mol dm−3 of Co4 (PW9 )2 and 5.0 × 10−4 mol dm−3
trolled by the GPES software. Voltammetric measurements were of MG, all in pH 4.0 acetate buffer. As the number of bilayers
performed at room temperature using a conventional three- increased there is an increment in the absorbance at max = 198,
electrode compartment cell. The working electrode was a glassy 257, 317, 426 and 653 nm, which is represented in the inset of
carbon electrode, GCE (3 mm diameter, BAS, MF-2012), the aux- Fig. 1, as straight lines in the plot of max vs. number of bilay-
iliary and reference electrodes were platinum wire (7.5 cm, BAS, ers. The electronic bands at max = 198 and 250 nm are assigned
MW-1032) and Ag/AgCl (sat. KCl) (BAS, MF-2052), respectively. The to Co4 (PW9 )2 , while the bands at max = 317, 426 and 653 nm
cell was enclosed in a grounded Faraday cage and kept under flow- to MG.
ing argon. A combined glass electrode (Crison) connected to a pH The linearity of Absmax vs. number of bilayers plots up to 15
meter Basic 20+ (Crison) was used for the pH measurements. bilayers suggests that the amount of material deposited in each
bilayer is constant and consistent with a stepwise and structurally
2.3. Film preparation regular growth process for the LbL hybrid. This is indicative of
strong interaction between the charged interlayers, contrasting
The {MG/Co4 (PW9 )2 }n multilayer films were deposited on the with the exponential signature of Absmax vs. number of film bilayers
following substrates: quartz slides for UV–vis spectroscopy, glass characteristic of weaker interlayer interactions.
42 H.C. Novais et al. / Applied Surface Science 347 (2015) 40–47

0.6
0.6 (a)
198 nm
257 nm , , 198 nm
0.5
0.6
0.5 317 nm , , 257 nm
426 nm , , 317 nm
0.4 653 nm , , 653 nm
0.5
0.4
Absorbance
Absorbance

0.3
0.4

Absorbance
0.3 0.2

0.1
0.3
0.2
0.0 0.2
0 2 4 6 8 10 12 14 16
0.1 Number of bilayers

0.1
0.0
200 400 600 800
0.0
Wavelength / nm 0 2 4 6 8 10 12 14 16
Number of bilayers
Fig. 1. UV–vis spectra of {MG/Co4 (PW9 )2 }n multilayers for n = 1–15 adsorbed
on a quartz slide (0.12 mol dm−3 of PEI, 1.0 × 10−3 mol dm−3 of Co4 (PW9 )2 and
5.0 × 10−4 mol dm−3 of MG, all in pH 4.0 acetate buffer). The inset shows the
(b)
absorbance at  = 198, 257, 317, 426 and 653 nm as a function of n. 0
0.7 , , 198
8 nm
, , 7 nm
257
0
0.6 , , 317
7 nm
Since the layer-by-layer self-assembly film growth can be , , 653
3 nm
influenced by some parameters, such as deposition time and com- 0
0.5
ponents concentration, two sets of experiments were conducted.
Absorbance

Fig. 2(a) depicts the plots of Absmax vs. number of bilay- 0

0.4
ers for PEI/Co4 (PW9 )2 /{MG/Co4 (PW9 )2 }n multilayers (n = 1–15)
at  = 198, 257, 317 and 653 nm for three different sets 0
0.3
of concentrations: (i) 1.0 × 10−3 mol dm−3 of Co4 (PW9 )2 and
5.0 × 10−4 mol dm−3 of MG; (ii) 5.0 × 10−3 mol dm−3 of Co4 (PW9 )2 0
0.2
and 5.0 × 10−4 mol dm−3 of MG and (iii) 1.0 × 10−3 mol dm−3 of
Co4 (PW9 )2 and 1.0 × 10−3 mol dm−3 of MG. In the three sets the PEI 0
0.1
concentration used was 0.12 mol dm−3 , all components solutions
were prepared in pH 4.0 acetate buffer solution and the deposition
0
0.0
time used was 10 min. In these conditions the films still continue to 0 2 4 6 8 10 1
12 14 16
growth linearly as the number of bilayers increases. These results N
Numbe r of biilayers
also show that when the Co4 (PW9 )2 concentration is increased 5
times no significant differences in absorbance values are observed. Fig. 2. Absorbance of {MG/Co4 (PW9 )2 }n multilayers (n = 1–15) at  = 198, 257,
317 and 653 nm for: (a) 0.12 mol dm−3 of PEI, 1.0 × 10−3 mol dm−3 of Co4 (PW9 )2
This behavior can be explained by the available positive sites for and 0.5 × 10−3 mol dm−3 of MG (); 0.12 mol dm−3 of PEI, 5.0 × 10−3 mol dm−3
electrostatic interactions: when the concentration of one of the of Co4 (PW9 )2 and 5.0 × 10−4 mol dm−3 of MG (); 0.12 mol dm−3 of PEI,
species is increased the absorbance may not reflect this because 1.0 × 10−3 mol dm−3 of Co4 (PW9 )2 and 1.0 × 10−3 mol dm−3 of MG ( ); (b)
the other species may not be in a sufficient concentration to allow 0.12 mol dm−3 of PEI, 1.0 × 10−3 mol dm−3 of Co4 (PW9 )2 and 0.5 × 10−3 mol dm−3
a higher film growth, since there is always an optimum correlation of MG, all in pH 4.0 acetate buffer for deposition times of 10 (), 5 () and 20 min
between the amount of positive and negative charges. When the ( ).
MG concentration is double no significant differences are observed
until n = 5; for 6 ≤ n ≤ 11 there is an increase in the Abs (≈25%) The surface coverage per layer,  UV , can be estimated
but above n = 11 the absorbance values reach a plateau. Conse- from the spectra of the multilayers, according to the Beer-
quently, the set of conditions selected for the subsequent studies Lambert law,  UV = A /2mε , where A is the absorbance at
was: 1.0 × 10−3 mol dm−3 of Co4 (PW9 )2 and 5.0 × 10−4 mol dm−3 the specified wavelength, m the number of layers and ε is
of MG. the isotropic molar absorption coefficient (M−1 cm−1 ). The
Fig. 2(b) shows the plots of Absmax vs. number of bilayers for factor 2 accounts for the fact that the film is covering both
PEI/Co4 (PW9 )2 /{MG/Co4 (PW9 )2 }n multilayers (n = 1–15) at  = 198, sides of the quartz slide. A set of eight solutions of each com-
257, 317 and 653 nm, using three depositions times (t = 5, 10 and pound (Co4 (PW9 )2 : 5.0 × 10−7 to 2.5 × 10−5 mol dm−3 and MG:
20 min). Typically, in the literature, an adsorption time between 1.0 × 10−6 to 4.0 × 10−5 mol dm−3 ) in acetate buffer (pH 4.0) was
5 and 20 min for each polyanion is used although, in most cases, used to determine the isotropic molar absorption coefficient:
the major amount of polyelectrolyte was in fact adsorbed within ε252 = 4.19 × 104 (Co4 (PW9 )2 ) and ε317 = 1.96 × 104 (MG) M−1 cm−1 ,
a few minutes [26]. Significant differences are observed in the leading to  UV252 = 0.269 ± 0.018 nmol cm−2 and
absorbance values for the three deposition times used but these  UV317 = 0.214 ± 0.044 nmol cm−2 .
are more evident for electronic band at  — — 198 nm. The increase
of t from 5 to 10 min leads to an increase in the Abs between 18% 3.2. SEM and XPS analysis
and 26% while for t = 10 min to t = 20 min, the increase is between
14% and 28%. Also, a more linear growth is observed for t val- Although UV–vis spectroscopy confirms the uniform growth
ues of 5 and 10 min (with 0.999 ≥ r ≥ 0.990) when compared with of the MG/Co4 (PW9 )2 multilayer assemblies via electrostatic
t = 20 min (0.983 ≥ r ≥ 0.961). Therefore, deposition time of 10 min LbL adsorption, this technique can only characterize the film
was selected for following studies. homogeneity as a whole. SEM can provide further information
 H.C. Novais et al. / Applied Surface Science 347 (2015) 40–47 43

Fig. 3. SEM images of {MG/Co4 (PW9 )2 }20 multilayers deposited on a glass slide at magnification 25,000× (a) and at magnification 150,000× (b).

concerning the surface morphology. Fig. 3 shows SEM images

4.0 (a) 2`
of {MG/Co4 (PW9 )2 }20 multilayer films deposited on a glass slide
at two different magnifications. Although the results show a
completely covered surface, the observed morphology consist of 2.0
1`
non-uniform distribution of dye molecules along with the POM
clusters. This roughened texture suggests that Co4 (PW9 )2 aggre- 0.0
gates to a certain level as the cationic dye molecules may reduce

i / µA
the Coulombic repulsion of adjacent POM centres.
-2.0 1
The presence of elements of each film component, methyl green
(C, N, Cl and Zn) and CoPOM (P, W, O and Co) also confirms the
successful fabrication of the hybrid {MG/Co4 (PW9 )2 }n films. The -4.0
presence of Cl and Zn in the XPS spectra of the films, although in
small atomic %, indicates that some counter cations from the methyl -6.0
green are incorporated within the films contributing to the overall 2
film charge compensation. Consequently, the film charge mech- -8.0
anism has two contributions: the intrinsic charge compensation, -1.0 -0.9 -0.8 -0.7 -0.6 -0.5 -0.4 -0.3 -0.2 -0.1
from the film components, which is the major contribution, and -3
extrinsic charge compensation from the counter cations from the E / V vs. Ag/AgCl (3 mol dm KCl)
methyl green (Tables 1 and 2).
The C 1s core level spectra (Fig. S2(a), SI) shows four decon- 7.0
voluted peaks at approximately 285.0, 286.4, 287.9 and 289.2 eV. (b)
slope = 1.27
The peak at 285 eV is attributed to the aliphatic carbon (–C–C– or
–C–H–), that at 286.4 eV to carbon bonded to the amine groups 6.0
(–C–NR2 , where R is C or H) and that at 287.9 to carbon-nitrogen
double bond (–C N) [40]. The peak at 289.2 eV can be ascribed 5.0 slope = 1.04
to residual organic impurities (CO2 and carbonate species). The ipc1
log or -log ip / A

appearance of two peaks in the N 1s spectra indicates the pres- ipa1

ence of two types of nitrogen atoms in the multilayer films. One of 4.0
ipc2
these peaks is attributed to the amine groups of MG (–NR2 ), corre- ipa2
sponding to the N 1s peak close to 400.1 eV; the second at 402.7 eV
-5.0 slope = 0.99
corresponds to protonated amine groups [40,41]. The Cl 2p spec-
trum can be deconvoluted into the 2p3/2 and 2p1/2 doublets at 200.6
and 202.2 eV, respectively, which are assigned to chloride anion. -6.0
The O 1s spectrum (Fig. S2(d), SI) was fitted with three peaks
slope = 1.26
at 530.9, 532.4 and 533.7 eV. The major peak at 530.9 eV can be
-7.0
ascribed to the O–W, that at 532.4 eV to O–P and the less intense at -1.6 -1.5 -1.4 -1.3 -1.2 -1.1 -1.0 -0.9 -0.8
533.7 eV to O–Co [42]. The XPS spectrum of the W 4f is shown in Fig- log v / V s
-1

ure S2(e). The binding energies of W4f can be resolved into 4f7/2 and
4f5/2 doublets caused by spin-orbital coupling at 35.7 and 37.9 eV. Fig. 4. Cyclic voltammograms of {MG/Co4 (PW9 )2 } bilayer in pH 4.0
Cobalt analysis was performed using 2p XPS high-resolution spec- CH3 COOH/CH3 COONa at different scan rates: 0.03, 0.04, 0.05, 0.06, 0.07, 0.08,
0.09, 0.1, 0.12 and 0.15 V s−1 (a); Plots of log ip vs. log  (b).
trum and two peaks were observed at 782.3 and 797.5 eV attributed
to 2p3/2 and 2p1/2 , respectively. However, for surface atomic % only
the 2p3/2 region was considered. The atomic % ratios of ≈1:2 for are shown in Fig. 4. Methyl green is not electroactive in the
Co:P and 1:9 for P:W confirm that the POM structure is maintained potential range studied so, the electrochemical responses are
in the films. attributed to the immobilized Co4 (PW9 )2 and correspond to W-
centred reduction processes (WVI → WV ) with E1/2 (1) = −0.658
3.3. Electrochemical characterization of {MG/Co4 (PW9 )2 }n films and E1/2 (2) = −0.843 vs. Ag/AgCl, scan rate 0.050 V s−1 . At pH 4.0
CH3 COOH/CH3 COONa buffer solution, the free POM (Figure S3, SI)
The cyclic voltammograms for a single {MG/Co4 (PW9 )2 } bilayer revealed one well defined W-centred reduction process with E1/2
obtained at different scan rates in pH 4.0 CH3 COOH/CH3 COONa, (2) = −0.820 V vs Ag/AgCl, which is similar to the corresponding
44 H.C. Novais et al. / Applied Surface Science 347 (2015) 40–47

Table 1
Core-level binding energies of each component for multilayer film of {MG/Co4 (PW9 )2 }20 obtained by curve fitting of XPS spectra and respective surface atomic percentages.

Binding energy (BE)a

Compound C 1s O 1s N 1s P 2p3/2 + 2p1/2 Cl 2p3/2 + 2p1/2 Co2p3/2 Zn2p3/2 W 4f7/2 + 4f5/2

285.0 (1.3) 530.9 (1.4) 400.1 (1.6) 134.0 (1.1) 200.6 (1.5) 782.4 (5.5) 1021.8 (3.3) 35.7 (1.2)
286.4 (1.3) 532.4 (1.4) 402.7 (1.6) 134.9 (1.1) 202.2 (1.5) 37.9 (1.2)
287.9 (1.3) 533.7 (1.4)
{MG/Co4 (PW9 )2 }20 film
289.2 (1.3)
Atomic %b
64.8 23.6 6.0 0.5 0.1 0.7 0.1 4.2
a
The values between brackets refer to the FWHM of the bands.
b
Determined by the areas of the respective bands in the high-resolution XPS spectra.

Table 2
Cyclic voltammetric data of [Fe(CN)6 ]3−/4− and [Ru(NH3 )6 ]3+/2+ complexes at bare GCE and modified with {MG/Co4 (PW9 )2 }n films,  = 0.1 V s−1 ; potentials units are in V vs.
Ag/AgCl.

n ipa (␮A) −ipc (␮A) Epa Epc |Ep | E1/2 a ipc (%)b

[Fe(CN)6 ]3−/4− 0 13.65 13.49 0.300 0.222 0.078 0.261 –

1 11.14 11.44 0.321 0.202 0.119 0.261 15.2
2 3.82 4.09 0.449 0.033 0.416 0.241 69.7
4 0.380 0.384 0.450 0.040 0.410 0.245 97.2

[Ru(NH3 )6 ]3+/2+ 0 12.45 12.46 −0.130 −0.195 0.065 −0.163 –

1 11.95 12.81 −0.134 −0.212 0.078 −0.173 −2.8
2 5.74 6.86 −0.150 −0.249 0.099 −0.200 44.9
3 4.03 4.83 −0.125 −0.241 0.116 −0.183 61.2
5 0.684 1.00 −0.124 −0.323 0.199 −0.224 92.0
a
E1/2 = E◦ = 0.5(Epa + Epc ).
b
ipc = [(ipc (GCE) − ipc (bilayer))/ipc (GCE)] × 100.

process in the immobilized Co4 (PW9 )2 ; the reduction process at 2`

4.0
less negative potential, that is observed in the immobilized POM at
E1/2 (1) = −0.658, is not well defined in the free POM. 2.0 1`
In the experimental timescale employed (scan rates in the
range 0.030–0.150 V s−1 ) the values of peak potential did not 0.0
change with scan rate. Fig. 4(b) depicts the plot of log ip vs log 
for {MG/Co4 (PW9 )2 } bilayer film. The cathodic and anodic peak
-2.0 1 5.0
ipc1
ipc2

currents of both W waves were directly proportional to the scan

i / µA

-4.0 4.0
rate (with 0.995 ≥ r ≥ 0.989), which indicates a surface-confined
process [43]. In addition, the anodic/cathodic peak-to-peak sep- -6.0 3.0
i / µA

aration (Ep ) for the tungsten waves was ca. 0.041–0.066 V.

The representative cyclic voltammograms as a function of the -8.0 2.0
2
number of layers for {MG/Co4 (PW9 )2 }n film up to 6 are depicted in
-10.0 1.0
Fig. 5. The cathodic and anodic peak potentials did not change with
an increase in the number of layers. The inset shows the plots of ipc1 -12.0 0.0
1 2 3 4 5 6
(for Epc1 = −0.688 V) and ipc2 (for Epc2 = −0.867 V) vs. the number of Number of layers
bilayers (n). From bilayers 1 to 5, the plots show linear growth, -14.0
indicating a uniform growth of the film for each deposition. Above -1.0 -0.9 -0.8 -0.7 -0.6 -0.5 -0.4 -0.3 -0.2 -0.1
these numbers of bilayers, peak currents progressive decrease and -3
E / V vs. Ag/AgCl (3 mol dm KCl)
the peaks become broad. This may suggest a change in the diffusion
regime, for example from a strictly surface-confined to a solution- Fig. 5. Cyclic voltammograms of {MG/Co4 (PW9 )2 }n multilayers for n = 0–6 in pH 4.0
type behaviour, with corresponding changes in the movement of CH3 COOH/CH3 COONa at 0.03 V s−1 . Inset shows ipc vs. number of layers.
charges and counter ions leading to an increased charge-transfer
resistance in the film structure.
Electrochemical surface coverages of electroactive species and  Epc2 = 0.155 ± 0.005 nmol cm−2 . The electrochemical surface
immobilized at electrode surfaces can be calculated from cyclic coverages obtained here are somewhat higher than those of other
voltammetry according to the equation G = (4ip RT )/(n2e F 2 nA), multilayer films based on POMs and polyelectrolytes [36,43,44].
where ip is the peak current (amperes), ne the number of These results show that the use of this molecular dye for film
electrons transferred,  the scan rate (V s−1 ), A the geometric assembly instead of the common polyelectrolytes leads to a higher
area of the electrode (0.0538 cm2 for GCE), R the gas con- amount of electrocatalysts available at the electrode surface which
stant, T the temperature (298 K) and F is Faraday’s constant constitutes an advantage when developing novel modified elec-
[43]. The POM surface coverage within the film was estimated trodes for applications in electrocatalysis and electroanalytical
using the peak current of Epc1 (one electron process) and Epc2 determinations. Also, these values are comparable to those esti-
(two electron process) for a single {MG/Co4 (PW9 )2 } bilayer. mated by UV–vis spectroscopy.
Peak potentials of Epc1 and Epc2 were plotted against scan rate Cyclic voltammetry was used to study film permeabil-
(0.030–0.150 V s−1 ) and the value of ip /v obtained was used to cal- ity properties of the multilayer films towards the negatively
culate the POM surface coverage,  Epc1 = 0.159 ± 0.020 nmol cm−2 charged [Fe(CN)6 ]3−/4− redox probe and the positively charged
 H.C. Novais et al. / Applied Surface Science 347 (2015) 40–47 45

15.0 films where polyelectrolytes are used [45]. The {MG/Co4 (PW9 )2 }n
3-
[Fe(CN)6] film also has a marked influence on redox probe voltammetric i–E
10.0 curve shape and peak potential separation. The peaks become broad
and progressively plateau-shaped with increasing number of bilay-
5.0 ers, suggesting a change in the diffusion regime from semi-infinite
linear diffusion to convergent diffusion to an array of sites behaving
as microelectrodes [46]. There is also an increase in peak poten-
0.0
tial separation (from 0.078 to 0.416 V), which is attributed to an
i / µA

increase in the film coverage that in turn increases the charge-

-5.0
transfer resistance of the probe redox process. Nevertheless, the
formal potential for the [Fe(CN)6 ]3−/4− couple does not change sig-
-10.0 nificantly, indicating that there is no measurable influence on the
GCE
underlying thermodynamics.
1
-15.0
2 Fig. 6(b) shows cyclic voltammograms of [Ru(NH3 )6 ]3+/2+ at the
4 electrode modified with {MG/Co4 (PW9 )2 }n for n = 0, 1, 2, 3 and
-20.0 5. At the bare electrode, the redox probe shows a well resolved
-0.2 0.0 0.2 0.4 0.6 0.8 redox process with Epc = −0.195 and Epa = −0.130 V, respectively,
E / V vs. Ag/AgCl (3 mol dm KCl)
-3
assigned to the RuIII/II reduction/oxidation. For an electrode coated
with a single bilayer, the cyclic voltammogram changes slightly
(Epc ≈ 0.013 V more negative, ipc increases ≈3%) but the probe still
15.0
(b) [Ru(NH3)6]
3+ undergoes electron transfer at the underlying electrode surface.
Increasing the number of bilayers to 2 leads to drastic changes in the
10.0 cyclic voltammograms, peak currents decrease approximately 45%
and peak potential separation increases from 0.065 to 0.090 V. From
5.0 n = 2 to 5, the peak currents still progressive decrease by 45–92% and
peak separation increases until 0.199 V. Although [Ru(NH3 )6 ]3+/2+
and last film layer (Co4 (PW9 )2 ) have opposite charges what would
0.0
i / µA

in principal promote electrostatic attraction, this is not observed.

The decrease in peak currents and increase in peak-to-peak separa-
-5.0 tion is probably due to the presence of fewer pathways, as the film
GCE becomes thicker which is confirmed by the higher electrochemical
1 surface coverages obtained when compared to those in the litera-
-10.0
2
ture for similar films [36,43]. These results indicate that not only the
3
5 electrostatic attractions or repulsions have effects on the kinetics
-15.0
of the redox reactions but also film thickness.
-0.6 -0.5 -0.4 -0.3 -0.2 -0.1 0.0 0.1 0.2 0.3
-3
3.4. Electrocatalytic properties of {MG/Co4 (PW9 )2 }n multilayer
E / V vs. Ag/AgCl (3 mol dm KCl) films

Fig. 6. Cyclic voltammograms of: (a) [Fe(CN)6 ]3−/4− (1 mM, 1 M KCl),  = 0.10 V s−1 ,
The applicability of POM species, as a reduction electrocatalyst,
at modified electrode with {MG/Co4 (PW9 )2 }n for n = 0, 1, 2 and 4; (b) [Ru(NH3 )6 ]3+/2+
(1 mM, 1 M KCl) for n = 0, 1, 2, 3 and 5. is based on the fact that their reduced forms, which are the catalytic
mediators, can provide a large number of electrons at a suitable, not
too negative, potentials. Heteropolyanions, in general, have been
[Ru(NH3 )6 ]3+/2+ . Electrostatic attraction between the multilayer proved to be excellent catalysts for the electroreduction of various
surface and the redox probe (in the case of opposite charges) should species [39,47,48] and have been extensively exploited. Nitrite and
facilitate the interfacial electron transfer process, whereas when iodate were chosen to test the electrocatalytic properties of the
they have the same charge (electrostatic repulsion) would make {MG/Co4 (PW9 )2 }n multilayer films.
the electron transfer reaction more difficult. Studies have shown Fig. 7(a) shows the CVs obtained for the {MG/Co4 (PW9 )2 }6
that the permeability towards these species can be tailored through modified electrode in the absence and in the presence of
the multilayer construction and deposition conditions [31] and that increasing concentrations of nitrite in the interval from 0 to
the number of multilayer films can influence the shape of the cyclic 1.50 × 10−4 mol dm−3 , at a scan rate 0.05 V s−1 . At pH 4.0, the peak
voltammograms [31,36,43]. at Epc2 = −0.850 V, that corresponds to the 2nd W-centred reduction
Fig. 6(a) shows cyclic voltammograms of [Fe(CN)6 ]3−/4 at the process (WVI →WV ), is enhanced by the addition of nitrite, whereas
electrode modified with {MG/Co4 (PW9 )2 }n for n = 0, 1, 2 and 4. the peak at Epc1 ≈ −0.666 V is almost unaffected. The direct elec-
At the bare GCE electrode, [Fe(CN)6 ]3−/4− shows the typical redox troreduction of nitrite/nitrous acid at the bare electrode, requires
process assigned to the FeIII/II reduction/oxidation of the probe at a large overpotential and no response is observed at a CGE in
the bare electrode with Epc = 0.222 and Epa = 0.300 V. Increasing the a solution containing NO2 − in the range of potentials used in
number of film bilayers, from 1 to 4, produces drastic and progres- this study. The increase of W-based cathodic peak current with
sive decrease in probe peak current by 15.2–97.2%. This shows that simultaneous decrease of the corresponding anodic peak current
the increase of bilayers number with a terminal negatively charged indicates that nitrite is being electrocatalytically reduced at the
[Co4 (H2 O)2 (PW9 )2 ]10− anion leads to a decrease in the number of modified electrode. The linear range for the catalytic current is
probe ions that reach the electrode substrate, due to electrostatic up to a nitrite concentration of 1.50 × 10−4 mol dm−3 , as shown in
repulsion of [Fe(CN)6 ]3−/4 by the negatively charged Co4 (PW9 )2 the inset, with a detection limit of 7.5 × 10−6 mol dm−3 (the detec-
anion and the presence of fewer pathways, as film becomes thicker. tion limit was calculated through the equation 3/slope, where 
Similar results were obtained for Eu(PMo11 )2 clusters [43]. This loss is the standard deviation of the blank). This value is much lower
of probe permeability is also similar to that observed at a number of than the 1.0 × 10−4 mol dm−3 for cellulose/PMo12 -based modified
46 H.C. Novais et al. / Applied Surface Science 347 (2015) 40–47

amount of electrocatalyst available for the electrocatalytic reaction

at the electrode surface that results from the use of molecular dye
methyl green as the positive component within the LBL multilayer
films.
Finally, electrocatalytic activity of POM towards oxidation or
reduction of substrates can be surveyed by analyzing the effect of
the substrates on the POM current intensity by using the following
equation [36]:
[ip(POM, substrate) − ip(POM) ]
CAT = 100% ×
ip(POM)

where ip(POM) and ip(POM , substrate) are the current intensities of

POMs in the absence and presence of the substrates, respec-
tively. The analysis of the corresponding current intensities at
{MG/Co4 (PW9 )2 }n for NO2 − and IO3 − lead to CAT = 114% and 192%,
respectively. For the free Co4 (PW9 )2 the CAT value obtained for
the electroreduction of IO3 − was 71%. These results show that
{MG/Co4 (PW9 )2 }n modified electrode behaves as a better electro-
catalyst for the reduction of iodate than for nitrite; the same can be
concluded for the electrocataltyic activity of free Co4 (PW9 )2 . More
importantly, as the free POM showed no electrocatalytic activity
towards NO2 – reduction and a CAT value for IO3 − reduction lower
than that of immobilized POM, it can be concluded that the immo-
bilization of Co4 (PW9 )2 in the LbL multilayered films was highly
advantageous for the application of Co4 (PW9 )2 in electrocatalysis.

4. Conclusions

Hybrid multilayer films composed of methyl green (MG) and

anionic sandwich-type polyoxometalate Co(PW9 )2 were success-
fully prepared on different substrates using the LbL self-assembly
method. Electronic spectroscopy revealed a uniform film grown,
which is typical of systems with strong interactions between the
component layers. SEM images revealed a completely covered sur-
face and XPS established the incorporation of both components in
the multilayer films.
Fig. 7. Cyclic voltammograms of {MG/Co4 (PW9 )2 }n in pH 4.0 buffer solution The multilayers films showed two surface-confined tungsten-
obtained in the absence and in the presence of added concentrations of: (a) nitrite: based reduction processes and film permeability, accessed using
(1) 0, (2) 10, (3) 20, (4) 35, (5) 50, (6) 75, (7) 100, (8) 125 and (9) 150 ␮mol dm−3 ; (b) [Fe(CN)6 ]3−/4− and [Ru(NH3 )6 ]3+/2+ , revealed that both electrostatic
iodate: (1) 0, (2) 10, (3) 50, (4) 90, (5) 120, (6) 150, (7) 175 and (8) 200 ␮mol dm−3 ; interactions and film thickness have effects on the kinetics of the
scan rate 0.050 V s−1 . The insets show the catalytic peak current vs. analyte concen-
redox probe reactions. Furthermore, the reduction of nitrite and
tration at −0.849 V.
iodate can be effectively electrocatalised by {MG/Co(PW9 )2 }n mul-
tilayer films with very low detection limits which arise from the
electrode [49] and 7.2 × 10−5 mol dm−3 for {PEI/SiW11 Co}7 /GCE use of the molecular dye methyl green as the positive component
[33]. No electrocatalytic activity was observed for Co4 (PW9 )2 in within the multilayer films.
solution towards nitrite reduction.
Fig. 7(b) depicts the CVs for the {MG/Co4 (PW9 )2 }6 modified
Acknowledgments
electrode in the absence and in the presence of increasing concen-
trations of iodate in the interval from 0 to 2.25 × 10−4 mol dm−3 ,
The authors thank Fundação para a Ciência e a Tecnologia
at scan rate 0.05 V s−1 . As observed for nitrite, the peak at
(FCT, Portugal) for financial support through projects PEst-
Epc2 = −0.850 V is significantly enhanced by the addition of iodate.
C/EQB/LA0006/2013, FCOMP-01-0124-FEDER-037285 and Opera-
The linear range for the catalytic current for iodate is up to
tion NORTE-07-0124-FEDER-000067–Nanochemistry. D.F. thanks
2.25 × 10−4 mol dm−3 , as shown in the insets and the detection
FCT for the fellowship SFRH/BPD/74877/2010. Thanks are also due
limit is 5.9 × 10−6 mol dm−3 . As for nitrite, the electrocatalytic
to COST Action CM-1203 PoCheMoN.
activity of Co4 (PW9 )2 in solution towards the reduction of iodate
was also studied. The results showed (Figure S4, SI) that the increase
in iodate concentration leads to an increase in the peak current at References
≈−0.744 V. The linear range is up to 1.0 × 10−4 mol dm−3 and the
[1] Y.F. Song, R. Tsunashima, Recent advances on polyoxometalate-based molecu-
detection limit is 9.1 × 10−6 mol dm−3 . lar and composite materials, Chem. Soc. Rev. 41 (2012) 7384–7402.
Although the linear range of concentrations is not very [2] J.M. Clemente-Juan, E. Coronado, A. Gaita-Arino, Magnetic polyoxometalates:
from molecular magnetism to molecular spintronics and quantum computing,
high for both analytes tested here, the limits of detection
Chem. Soc. Rev. 41 (2012) 7464–7478.
are very good and lower than the 2.95 × 10−5 mol dm−3 for [3] D.W. Fan, J.C. Hao, Q. Wei, Assembly of polyoxometalate-based composite
{Os-poly/Eu(PMo11 )2 }6 /GCE [36] and 5.8 × 10−5 mol dm−3 for materials, J. Inorg. Organomet. Polym. Mater. 22 (2012) 301–306.
{PEI/SiW11 Co}7 /GCE [33]. The very low detection limits obtained [4] D.E. Katsoulis, A survey of applications of polyoxometalates, Chem. Rev. 98
(1998) 359–387.
with {MG/Co4 (PW9 )2 }n multilayer films for both the reduction [5] M. Sosnowska, M. Goral-Kurbiel, M. Skunik-Nuckowska, R. Jurczakowski,
of NO2 − and IO3 − is a straightforward consequence of the larger P.J. Kulesza, Hybrid materials utilizing polyelectrolyte-derivatized carbon
 H.C. Novais et al. / Applied Surface Science 347 (2015) 40–47 47

nanotubes and vanadium-mixed addenda heteropolytungstate for efficient [27] G. Decher, M. Eckle, J. Schmitt, B. Struth, Layer-by-layer assembled multicom-
electrochemical charging and electrocatalysis, J. Solid State Electrochem. 17 posite films, Curr. Opin. Colloid Interface Sci. 3 (1998) 32–39.
(2013) 1631–1640. [28] B. Wang, R.N. Vyas, S. Shaik, Preparation parameter development for layer-
[6] X. Zhang, Q. Chen, D.C. Duncan, R.J. Lachicotte, C.L. Hill, Multiiron polyoxoan- by-layer assembly of keggin-type polyoxometalates, Langmuir 23 (2007)
ions. Synthesis, characterization, X-ray crystal structure, and catalytic H2O2- 11120–11126.
based alkene oxidation by (n-C4 H9 )(4)N (6) Fe-4(III)(H2 O)(2)(PW9 O34 )(2), [29] F. Caruso, D.G. Kurth, D. Volkmer, M.J. Koop, A. Muller, Ultrathin molybde-
Inorg. Chem. 36 (1997) 4381–4386. num polyoxometalate-polyelectrolyte multilayer films, Langmuir 14 (1998)
[7] C.L. Hill, C.M. Prossermccartha, Homogeneous catalysis by transition-metal 3462–3465.
oxygen anion clusters, Coord. Chem. Rev. 143 (1995) 407–455. [30] S.Y. Gao, T.H. Li, X. Li, R. Cao, Electrochemical behavior and mul-
[8] J.A.F. Gamelas, M.S. Balula, H.M. Carapuca, A.M.V. Cavaleiro, Unusual elec- tilayer films of the sandwich-type polyoxotungstate complex
trochemical reduction of copper(II) to copper(I) in polyoxotungstates, {K10Co4(H2O)(2)(PW9O34)(2)}, Mater. Lett. 60 (2006) 3622–3626.
Electrochem. Commun. 5 (2003) 378–382. [31] S.Q. Liu, D.G. Kurth, B. Bredenkotter, D. Volkmer, The structure of self-assembled
[9] M.S. Balula, J.A. Gamelas, H.M. Carapuca, A.M.V. Cavaleiro, W. Schlindwein, multilayers with polyoxometalate nanoclusters, J. Am. Chem. Soc. 124 (2002)
Electrochemical behaviour of first row transition metal substituted polyox- 12279–12287.
otungstates: A comparative study in acetonitrile, Eur. J. Inorg. Chem. (2004) [32] D.M. Fernandes, H.M. Carapuca, C.M.A. Brett, A.M.V. Cavaleiro, Electro-
619–628. chemical behaviour of self-assembly multilayer films based on iron-
[10] X.Y. Zhang, G.B. Jameson, C.J. Oconnor, M.T. Pope, High-valent manganese in substituted alpha-Keggin polyoxotungstates, Thin Solid Films 518 (2010)
polyoxotungstates. 2. Oxidation of the tetramanganese heteropolyanion Mn- 5881–5888.
4(H2O)(2)(PW9O34)(2) (10), Polyhedron 15 (1996) 917–922. [33] D.M. Fernandes, C.M.A. Brett, A.M.V. Cavaleiro, Layer-by-layer self-assembly
[11] H.J. Lv, J. Song, Y.V. Geletii, J.W. Vickers, J.M. Sumliner, D.G. Musaev, P. Koger- and electrocatalytic properties of poly(ethylenimine)-silicotungstate
ler, P.F. Zhuk, J. Bacsa, G.B. Zhu, C.L. Hill, An exceptionally fast homogeneous multilayer composite films, J. Solid State Electrochem. 15 (2011)
carbon-free cobalt-based water oxidation catalyst, J. Am. Chem. Soc. 136 (2014) 811–819.
9268–9271. [34] S.P. Liu, L. Xu, F.Y. Li, W.H. Guo, Y. Xing, Z.X. Sun, Carbon nanotubes-assisted
[12] J.J. Stracke, R.G. Finke, Electrocatalytic water oxidation beginning with the polyoxometalate nanocomposite film with enhanced electrochromic perfor-
cobalt polyoxometalate Co-4(H2O)(2)(PW9O34)(2) (10-): identification of het- mance, Electrochim. Acta 56 (2011) 8156–8162.
erogeneous CoOx as the dominant catalyst, J. Am. Chem. Soc. 133 (2011) [35] Y. Lan, E.B. Wang, Y.H. Song, Y.L. Song, Z.H. Kang, L. Xu, Z. Li, An effective
14872–14875. layer-by-layer adsorption and polymerization method to the fabrication of
[13] Q.S. Yin, J.M. Tan, C. Besson, Y.V. Geletii, D.G. Musaev, A.E. Kuznetsov, Z. Luo, polyoxometalate-polypyrrole nanoparticle ultrathin films, Polymer 47 (2006)
K.I. Hardcastle, C.L. Hill, A fast soluble carbon-free molecular water oxidation 1480–1485.
catalyst based on abundant metals, Science 328 (2010) 342–345. [36] D.M. Fernandes, J.G. Vos, C. Freire, Europium phosphomolybdate and osmium
[14] Z.Q. Huang, Z. Luo, Y.V. Geletii, J.W. Vickers, Q.S. Yin, D. Wu, Y. Hou, Y. Ding, J. metallopolymer multi-functional LbL films: Redox and electrocatalytic prop-
Song, D.G. Musaev, C.L. Hill, T.Q. Lian, Efficient light-driven carbon-free cobalt- erties, J. Colloid Interface Sci. 420 (2014) 127–135.
based molecular catalyst for water oxidation, J. Am. Chem. Soc. 133 (2011) [37] L. Ning, W.J. Zhang, H. Yan, H.J. Pang, H.Y. Ma, Y. Yu, An electrochromic com-
2068–2071. posite film of Preyssler-type phosphotungstate decorated by AuNPs, J. Colloid
[15] J. Wu, L.W. Liao, W.S. Yan, Y. Xue, Y.F. Sun, X. Yan, Y.X. Chen, Y. Xie, Polyoxome- Interface Sci. 403 (2013) 91–98.
talates immobilized in ordered mesoporous carbon nitride as highly efficient [38] H.Y. Ma, Z.J. Zhang, H.J. Pang, S. Li, Y.Y. Chen, W.J. Zhang, Fabrication and elec-
water oxidation catalysts, ChemSusChem 5 (2012) 1207–1212. trochemical sensing property of a composite film based on a polyoxometalate
[16] V. Ball, F. Bernsmann, S. Werner, J.C. Voegel, L.F. Piedra-Garza, U. Kortz, Polyoxo- and palladium nanoparticles, Electrochim. Acta 69 (2012) 379–383.
metalates in polyelectrolyte multilayer films: direct loading of H7P8W48O184 [39] M.T. Pope, Heteropoly and Isopoly Oxometalates, Spring Verlag, 1983.
(33-) vs. Diffusion into the film, Eur. J. Inorg. Chem. (2009) 5115–5124. [40] M. Finsgar, S. Fassbender, S. Hirth, I. Milosev, Electrochemical and XPS study
[17] D.G. Kurth, D. Volkmer, R.V. Klitzing, Multilayer Thin Films: Sequential Assem- of polyethyleneimines of different molecular sizes as corrosion inhibitors for
bly of Nanocomposite Materials, 2nd ed., Wiley-VCH, New York, 2012. AISI 430 stainless steel in near-neutral chloride media, Mater. Chem. Phys. 116
[18] A. Proust, B. Matt, R. Villanneau, G. Guillemot, P. Gouzerh, G. Izzet, Function- (2009) 198–206.
alization and post-functionalization: a step towards polyoxometalate-based [41] L. Zhang, Y.H. Shen, A.J. Xie, S.K. Li, C. Wang, One-step synthesis of silver
materials, Chem. Soc. Rev. 41 (2012) 7605–7622. nanoparticles in self-assembled multilayered films based on a Keggin structure
[19] S. Polarz, B. Smarsly, C. Goltner, M. Antonietti, The interplay of colloidal organi- compound, J. Mater. Chem. 18 (2008) 1196–1203.
zation and oxo-cluster chemistry: polyoxometalate-silica hybrids—materials [42] P. Perez-Romo, C. Potvin, J.M. Manoli, M.M. Chehimi, G. Djega-Mariadassou,
with a nanochemical function, Adv. Mater. 12 (2000) 1503–1507. Phosphorus-doped molybdenum oxynitrides and oxygen-modified molybde-
[20] D.M. Fernandes, C.M.A. Brett, A.M.V. Cavaleiro, Preparation and electrochem- num carbides: synthesis, characterization, and determination of turnover rates
ical properties of modified electrodes with Keggin-type silicotungstates and for propene hydrogenation, J. Catal. 208 (2002) 187–196.
PEDOT, J. Electroanal. Chem. 660 (2011) 50–56. [43] D.M. Fernandes, C. Freire, Hybrid photochromic multilayer films based on
[21] D.M. Fernandes, C.M.A. Brett, A.M.V. Cavaleiro, Modified electrodes with chitosan and europium phosphomolybdate, J. Appl. Electrochem. 44 (2014)
Keggin-type silicotungstates and poly(brilliant cresyl blue), J. Solid State Elec- 655–665.
trochem. 16 (2012) 2267–2273. [44] D.M. Fernandes, D. Juliao, C. Pereira, D. Ananias, S.S. Balula, C. Freire,
[22] M. Clemente-Leon, E. Coronado, C.L. Gomez-Garcia, C. Mingotaud, S. Ravaine, Hybrid layer-by-layer films based on lanthanide-bridged silicotungstates
G. Romualdo-Torres, P. Delhaes, Polyoxometalate monolayers in Langmuir- and poly(ethylenimine), Colloid Surf. A-Physicochem. Eng. Asp. 415 (2012)
Blodgett films, Chem. Eur. J. 11 (2005) 3979–3987. 302–309.
[23] P.J. Kulesza, M. Chojak, K. Miecznikowski, A. Lewera, M.A. Malik, A. Kuhn, [45] N. Gu, D. Wei, L. Niu, A. Ivaska, Study on charge transfer reactions at multilayers
Polyoxometallates as inorganic templates for monolayers and multilayers of of polyoxometalates clusters and poly(allylamine hydrochloride) (Grotthuss-
ultrathin polyaniline, Electrochem. Commun. 4 (2002) 510–515. 018), Electrochim. Acta 51 (2006) 6038–6044.
[24] L. Adamczyk, P.J. Kulesza, K. Miecnikowski, B. Palys, M. Chojak, D. Krawczyk, [46] A.J. Bard, L.R. Faulkner, Electrochemical Methods, Fundamentals and Applica-
Effective charge transport in poly(3,4-ethylenedioxythiophene) based hybrid tions, New York, 2001.
films containing polyoxometallate redox centers, J. Electrochem. Soc. 152 [47] M. Sadakane, E. Steckhan, Electrochemical properties of polyoxometalates as
(2005) E98–E103. electrocatalysts, Chem. Rev. 98 (1998) 219–237.
[25] A. Kuhn, F.C. Anson, Adsorption of monolayers of P2Mo18O626- and deposi- [48] S.Q. Liu, Z.Y. Tang, Polyoxometalate-based functional nanostructured
tion of multiple layers of Os(bpy)(3)(2 + )-P2Mo18O626- on electrode surfaces, films: current progress and future prospects, Nano Today 5 (2010)
Langmuir 12 (1996) 5481–5488. 267–281.
[26] P. Bertrand, A. Jonas, A. Laschewsky, R. Legras, Ultrathin polymer coatings by [49] Y. Liang, P. He, Y.J. Ma, Y. Zhou, C.H. Pei, X.B. Li, A novel bacterial cellulose-
complexation of polyelectrolytes at interfaces: suitable materials, structure and based carbon paste electrode and its polyoxometalate-modified properties,
properties, Macromol. Rapid Commun. 21 (2000) 319–348. Electrochem. Commun. 11 (2009) 1018–1021.