This is the peer reviewed version of the following article: ChemElectroChem 2019, 6, 2282–2289, which has been
published in final form at https://chemistry-europe.onlinelibrary.wiley.com/doi/10.1002/celc.201900218. This article
may be used for non-commercial purposes in accordance with Wiley Terms and Conditions for Self-Archiving.

Fluorine-Doped Tin Oxide/Alumina as Long-Term Robust
Conducting Support for Earth-Abundant Water Oxidation
Electrocatalysts
Felipe A. Garcés-Pineda, Jesús González-Cobos, Mabel Torrens, and José R. GalánMascarós
In this work we apply a novel and simple methodology to
develop a composite that consists of anodic aluminum oxide
(AAO) covered by a fluorine-doped tin oxide (FTO) film. The
composite presents suitable morphological and electrical properties to be used as support for water electrooxidation catalysts
based on non-noble metals. AAO substrates were decorated
with FTO via spray pyrolysis, followed by deposition of catalyst
nanoparticles consisting of a cobalt-analogue of Prussian Blue
that promotes the oxygen evolution reaction. The electrodes

were characterized by X-ray Diffraction, Environmental Scanning
Electron Microscopy and resistivity measurements. A promising
electrocatalytic activity of the optimized AAO/FTO/PB electrode
was observed at neutral pH and ambient conditions. The
obtained results open a feasible strategy in the search for
competitive water oxidation electrodes based on the combination of earth-abundant metal catalysts on FTO-covered anodized alumina supports.

1. Introduction

reactions such as electroreduction of chloroform,[7] hydrogen
evolution reaction,[8] electrooxidation of hydrazine,[9] urea,[10]
methanol[11–12] or ethanol.[13] However, since AAO is electrically
insulator, it is not common the use of this material as
permanent support for electrocatalysts and, in general, for
applications where high electrical conductivity is required. In
these cases, AAO needs to be appropriately assembled with a
conductive component. For instance, Altuntas and Buyukserin
obtained an AAO modified with carbon to develop a biomaterial nerve tissue,[14] and Wierzbicka and Sulka used an Aucovered AAO as sensitive electrode for electrochemical determination of epinephrine.[15]
In the present study our interest is focused on the
application of AAO in water electrolysis, since one of the main
bottlenecks in this field is the search for stable electrodes able
to survive to challenging redox and pH conditions.[16–18] Noble
metal electrodes are traditionally used for water oxidation,
specially at low pH.[19–20] Ni-based electrodes are preferred at
high alkaline pH[8,21] but they rapidly get oxidized when the pH
is lowered. Carbon electrodes are pH-stable but unstable during
oxygen evolution catalysis.[22] In this sense, AAO can be an
alternative. This material has been recently used as support of
Ir[23–24] and Pt[24] catalysts for water oxidation in acidic media,
demonstrating the stability of AAO-supported catalysts under
electrooxidation conditions. Other current bottleneck in the
development of water electrolysis technology is precisely the
substitution of this kind of noble, expensive catalysts. Increasing
efforts are being made towards the implementation of Earthabundant transition-metal oxides.[18,25–26] In this context, a
cobalt-analogue of Prussian Blue (PB), cobalt hexacyanoferrate
(CoHCF), has been reported as a good candidate as water
oxidation catalyst showing an efficient and robust activity.[27–32]
Herein we aim to use anodic aluminum oxide as support for
Prussian Blue nanoparticles. However, AAO/PB electrodes could

Porous anodic aluminum oxide (AAO) has been extensively
studied for the last years in different fields in nanoscience and
nanotechnology.[1–2] AAO is a honeycomb-like porous material
formed by electrochemical anodization of aluminum in an acid
electrolyte. As a result, an ordered nanoporous structure with
very high surface area is obtained. Moreover, this material is
mechanically, chemically and thermally stable. AAO hardness,
corrosion resistance and erosion resistance have shown to be
improved after the aluminum anodization process.[3] For
instance, AAO is attractive as template or host material for the
growth of ordered nanotube or nanowire arrays for electrochemical applications like energy storage and conversion.[4] As a
few examples, AAO templates have been used to fabricate a
nanostructured CdS-CdTe solar cell[5] or a photoanode for water
splitting based on Fe2TiO5 nanotubes.[6] In these cases the AAO
support is removed after the deposition of the guest material
leading to free-standing nanoporous structures. The AAO
removal is typically performed by immersing the composite
material into a concentrated NaOH or H3PO4 solution. There are
several examples of nanoporous electrodes obtained in this
way which have been tested for different electrocatalytic
]
Dr. F. A. Garcés-Pineda, Dr. J. González-Cobos, Dr. M. Torrens,
Prof. J. R. Galán-Mascarós
Institute of Chemical Research of Catalonia (ICIQ)
The Barcelona Institute of Science and Technology (BIST)
Av. Països Catalans, 16, Tarragona, 43007 Spain
E-mail: fgarces@iciq.es
jcobos@iciq.es
Prof. J. R. Galán-Mascarós
ICREA
Passeig Lluis Companys, 23, Barcelona, 08010 Spain
Supporting information for this article is available on the WWW under
https://doi.org/10.1002/celc.201900218

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57

not be directly used as anodes for water splitting given the
electrically insulating character of this hypothetical composite.
Our strategy in the present work consists of a controlled
deposition of a conductive material over the AAO surface
before the PB addition. SnO2 is one of the most studied
members in the family of transparent conductive oxides (TCO)
and, particularly, F-doped tin oxide (FTO).[6,33–36] Besides its
transparency in the visible region of the electromagnetic
spectrum and low electrical resistance, it can be easily
deposited on different surfaces by spray pyrolysis.[35–36] We
propose for the first time in literature the use of FTO-covered
AAO as water oxidation catalyst support. In this way, it is
expected that FTO provides a suitable electrical conductivity on
the surface and that AAO induces an increase in the rugosity
and number of defects of the FTO layer. AAO/FTO composites
with different FTO layer thicknesses have been characterized in
terms of surface morphology and electrical properties, and
CoHCF nanoparticles supported on the optimum AAO/FTO
substrate have exhibited a promising performance for water
splitting at neutral pH. The results shown below indicate that
AAO/FTO is a viable, efficient and long-term robust electrode to
support non-noble water oxidation catalysts.

2. Results and Discussion
Porous alumina substrates were modified by depositing fluorine-doped tin oxide (FTO) layers by spray pyrolysis. SnO2:F
coverage was controlled by varying the deposition time.
Figure 1 shows the scanning electron microscopy (ESEM)

Figure 1. ESEM images from two different layers of AAO/FTO template. On
the left, an AAO substrate with a FTO layer obtained after 10 minutes of
deposition time. On the right, an image with the FTO deposited for 120
minutes.

images for AAO/FTO samples after 10 and 120 min deposition.
One can observe that AAO surface is covered by FTO nanograins and both the number and the size of these grains show
to be increased along the deposition time, finally obtaining a
fairly uniform layer composed of FTO grains with diameters of
the order of μm. A similar FTO grain dependence on deposition
time was recently shown by other authors on glass substrate.[37]
The AAO/FTO pore size is also reduced along FTO deposition, as it can be observed in Figure 2. The bare AAO template
used herein presents an average pore size of c.a. 220 nm and

Figure 2. ESEM images in top view from samples obtained after different
deposition times of FTO onto AAO templates.

finally, after 120 min, the AAO outermost surface seems to be
almost fully covered by FTO without apparent open pores
(maximum pore size of c.a. 50 nm). This effect may be linked to
the grain nucleation. For longer times (i. e. thicker FTO layers)
the adatoms coalescence increases and this directly affects the
crystal growth. In the AAO/FTO sample prepared for 120
minutes it can also be appreciated a high roughness on the
membrane surface.
A partial penetration of FTO into the alumina pores is
confirmed from the cross-section image shown in Figure 3,
where several EDX spectra were taken at different depths in the
AAO. Although FTO signal decreased along the pore depth, it
can be observed that FTO penetration takes place at least
20 μm in depth. Despite this, on the basis of these micrographs
one can anticipate that the FTO coverage is mostly produced
on the AAO outermost surface in contrast, for example, to a
previous study where a nanotubular electrode was obtained by
introducing FTO into the channels of an AAO substrate.[38] It was
possible in this case due to a larger initial pore diameter of
AAO, c.a. 900 nm (vs. c.a. 220 nm in the present work). Despite
the high specific surface area of the bare AAO substrate
employed herein, the pore diameter is not wide enough to
allow the full covering in depth by FTO during the spray
pyrolysis deposition. Thus the AAO/FTO/PB electrodes tested
below probably present lower active surface area than other

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57

ICSD database) which are shown as vertical dotted lines in
Figure 4a. This indicates the presence of the tetragonal phase of
SnO2 with the characteristic crystalline directions such as [1 0 1],
[2 0 0] and [2 1 1]. The Full Width at Half Maximum (FHWM)
parameter decreases for longer deposition times which indicates an increase of the FTO crystallite size. This is confirmed in
Figure 4b, where the values calculated from the Scherrer
equation are shown. The crystallite size increases from c.a. 8 to
c.a. 19 nm along the deposition time.
Electrical characterization of the AAO/FTO samples was also
performed by means of a method based on I-V characteristics
obtained in a planar configuration (see inset of Figure 5) which

Figure 3. ESEM image in cross section of the AAO/FTO layer deposited for
120 min along with EDX spectra obtained at a) c.a. 1 μm, b) c.a. 10 μm and c)
c.a. 20 μm in depth.

electrodes that could be prepared by using AAO supports with
larger initial pore diameter.
Structural analysis of the samples by X-ray diffraction (XRD)
was also carried out. The XRD patterns (Figure 4a) reveal the
polycrystalline nature of the FTO films with no preferential
orientation. The diffraction peaks are coincident with those of
the Cassiterite phase (reference 1-072-1147, reported in the

Figure 4. a) X-ray diffraction patterns for different deposition times of FTO
onto AAO membranes. The dotted lines correspond to the reference 1-0721147 reported in the ICSD database. b) Crystallite size dependence on
deposition time.

Figure 5. a) I-V characteristic for two AAO/FTO layers after different
deposition times. b) Electrode resistivity decrease with the FTO layer
thickness increase. The inset of this figure shows the setup configuration
employed for these measurements.

has also been applied in other works.[14] In Figure 5a it can be
observed the I-V curve for the different AAO/FTO composites,
from whose slope the sheet resistance can be obtained. First
this value decreased sharply from 3.24 × 107 Ω/sq (on the FTO
layer deposited for 10 min) to 708 Ω/sq (on the one deposited
for 60 min) and, from then, it continued slightly decreasing up
to reach 330 Ω/sq (on the FTO layer deposited for 120 min).
This decrease in the electrical resistance can be attributed to
the influence of the deposition time in the achievement of
larger FTO grains and a more intimate contact between grains
which may improve the carrier transport. The higher the grain
size and the higher the grain connectivity, the lower the sheet
resistance, as it has also been observed by other authors on
FTO layers deposited on glass substrates.[35,37,39] Figure 5b shows
the evolution of the resistivity along the FTO deposition time.
This parameter was estimated from the sheet resistance values
and the corresponding thickness of the FTO layers. The thickness increased during the deposition procedure up to reach
2.1 μm onto each AAO side after 120 min and the growth rate
of the FTO layer was estimated at c.a. 35 nm min-1. One can
observe a sharp decrease of resistivity of four orders of
magnitude, to 0.149 Ω cm, for the first 60 min of FTO deposition. The decrease of the resistivity with the increase of the FTO

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57

thickness has also been reported in other works.[37,40] However,
in contrast to the sheet resistance trend, the resistivity remained
fairly constant from this moment and it was finally measured at
0.138 Ω cm after 120 min.
CoHCF nanoparticles were deposited onto the different
AAO/FTO supports following the same dip-coating procedure in
each case, as described in the experimental section. Figure 6

Figure 6. (a) EDX spectra and (b) XRD diffractograms from AAO/FTO/CoHCF
electrode (120 min, 330 Ω/sq) both before and after the electrocatalytic
tests.

shows the XRD and EDX spectra of the AAO/FTO/PB electrode
(with FTO layer deposited for 120 min) and the AAO/FTO (blank)
electrode, both in the fresh state and after being used for the
electrocatalytic tests. In the EDX spectra (Figure 6a) the
incorporation of Fe and Co atoms clearly appears in the PBdecorated AAO/FTO electrode. In the XRD diffractograms of the
AAO/FTO sample (Figure 6b, left) the same six FTO diffraction
peaks that were observed in Figure 4a are identified. In the case
of the AAO/FTO/PB sample (Figure 6b, right), four new
diffraction peaks are identified at 2θ = 17.9°, 25.4°, 36.3° and
58°, which can be attributed to Prussian blue and are characteristic of a face centered cubic structure.[41] Similar EDX and XRD
patterns related to FTO and Prussian blue are found before and
after the electrocatalytic measurements. This denotes that no
significant catalyst leaching or change in phase took place,
which confirms the stability of both the active electrode and
the blank under the studied electrochemical conditions. Moreover, PB crystallite sizes of 15.7 nm and 14.5 nm are calculated
from the diffractograms of the AAO/FTO/PB electrode before
and after the electrocatalytic tests, respectively. This size is
lower than that previously reported for other PB-based electrodes prepared by different deposition methods, i. e. dropcasting,[32] spray deposition,[31] electrodeposition[29] or a dipcoating procedure similar to the one employed herein.[30] In the

present case the most influential feature seems to be the
surface rugosity and high number of defects generated onto
the AAO/FTO substrate which may lead to a large nanostructuring that allows the formation of such small PB
crystallites. ESEM images were also obtained both before and
after the electrocatalytic tests (see Figure S1 in Supplementary
Information) and no apparent changes in electrode morphology
can be appreciated.
Figure 7 shows the linear sweep voltammetry (LSV) performed in 50 mM potassium phosphate (KPi) electrolyte at

Figure 7. Linear sweep voltammetry for AAO/FTO/CoHCF electrodes with
different FTO deposition times in 50 mM KPi electrolyte at pH 7.0.

pH 7.0 with the four AAO/FTO/PB electrodes with the FTO layers
that showed lower resistivity values in Figure 5, i. e. those
deposited for 60, 80, 100 and 120 min, in order to compare
their electrocatalytic activity. The composite materials with FTO
layers deposited for 60 and 80 min show negligible activity. The
electrode with 100 min of FTO deposition shows a significant
current density at the higher potentials, but the oxygen
evolution curve is not well defined yet. The most active
electrode is, by far, the one containing the FTO layer deposited
for longest time, which reaches c.a. 32 mA cm-2 at 2.4 V vs. RHE.
Taking into account that Prussian Blue catalyst was deposited
for the same dip-coating time and that the obtained resistivity
was the same in all these cases, one of the main reasons for the
increase of the electrocatalytic activity with the FTO deposition
time could be the surface roughness enhancement. The roughness has been typically found to increase with the thickness in
studies on FTO films deposition by different methods[37,40, 42–43]
and this feature can lead to an increase of the number of active
Prussian Blue nanoparticles. Moreover, as the FTO deposition
time is shorter, it is also plausible that these nanoparticles were
deposited to a greater extent into the pores of the AAO/FTO
composite. Although the specific surface area of the AAO
support and the blocking extent of the pores by FTO were not
quantified, the pore diameter left after FTO spray pyrolysis
clearly decreases along deposition time, as observed in Figure 2.
Hence, the electrodes with FTO layers deposited for less than

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57

120 min likely lose most of the electrocatalytic activity from the
CoHCF nanoparticles placed inside the AAO pores due to strong
mass transfer limitations. In order to avoid mass transfer
limitations in these cases different AAO substrates should be
employed with higher initial pore diameter than that used in
the present work (c.a. 200 nm).
Focusing on the case of the AAO/FTO/PB electrode with the
FTO layer deposited for 120 min, one can clearly observe the
typical exponential behavior due to water oxidation process
after 2 V vs. RHE. Before this potential, a pre-oxidation peak is
present at around 1.9 V vs. RHE. A similar peak is typically
observed during LSV performed in other studies on water
electrolysis with PB catalysts[28–29,32] and it can be attributed to
an electron transfer process due to the CoIIFeII-CoIIIFeII couple.
As it can be observed in Figure 8, the current density values

Figure 9. Tafel plot of the steady-state current density of the AAO/FTO/
CoHCF and AAO/FTO electrodes in 50 mM KPi electrolyte at pH 7.0. The
linear region shows the different Tafel slope for both electrodes.

Figure 8. Linear sweep voltammetry for AAO/FTO/CoHCF electrode (120 min,
330 Ω/sq) in 50 mM KPi electrolyte at pH 7.0 (red line). Black line represents
the blank substrate, AAO/FTO (i. e. without CoHCF).

obtained with the AAO/FTO/CoHCF electrode show to be much
higher than those obtained with the AAO/FTO blank electrode,
specially at higher potentials. This denotes that the electrocatalytic activity fairly proceeds from the PB nanoparticles.
In the Tafel plot (Figure 9) it can be observed a linear
relationship between log j and the overpotential (η) in an η
range of 0.3-0.5 V and 0.5-0.75 V for the AAO/FTO/PB and the
AAO/FTO electrodes, respectively. In these regions the Tafel
slope decreases from 99 mV/dec (blank electrode) down to
82 mV/dec for the PB-decorated substrate, which proves the
catalytic process. In this kind of Prussian Blue-analogues the
water oxidation catalytic activity is typically assigned to the Co
centers while the function of the Fe centers is rather structural
and
electronic
(enhancing stability
and
electron
conduction).[28,32] It was previously demonstrated the competitiveness of PB-based electrodes in kinetic terms with respect to
CoOx catalysts[29] and, very interestingly, the Tafel slope
obtained with the AAO/FTO/PB electrode is even lower than
the values previously shown at pH 7 by Aksoy et al.,[27] Alsaç
et al.[28] and Pintado et al.[29] with other PB-based electrodes

prepared on flat substrates. The obtained electrocatalytic
activity may be favored by the small PB crystallite size
mentioned above but, specially, by the particular surface
morphology of the AAO/FTO support. The use of this kind of
porous support and the resultant roughness of the FTO
conductive layer likely favor the presence of a higher amount of
catalytic active sites leading to an improved electrocatalytic
activity.
In order to check the long-term stability of the AAO/FTO/PB
catalyst, a chronoamperometry (CA) was carried out for
100 hours under the same experimental conditions in the
oxygen evolution voltage region, at 2.2 V vs. RHE (Figure 10).
With both the AAO/FTO/PB and the blank electrode a sharp
decrease of the current density is observed at the beginning of
the CA, which is mainly due to the fast decay of capacitive
current. The AAO/FTO sample showed a negligible activity

Figure 10. Bulk electrolysis of the AAO/FTO/CoHCF electrode at 2.2 V vs. RHE
in 50 mM KPi electrolyte at pH 7.0, along with the blank measurement.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57

under steady-state conditions. In the case of the AAO/FTO/PB
electrode, a slow decrease of current density is observed. One
cannot totally discard some instability of the outermost FTO
crystallites since the blank data shows some minor deactivation.
However, this decay is certainly far away from the current
decrease on the PB-decorated electrode, in agreement with the
good FTO chemical stability reported under different pH and
potential conditions, compared with other TCOs like indium tin
oxide (ITO) or antimony-doped tin oxide (ATO).[44–45] Therefore,
the current decay of the AAO/FTOCoHCF sample during the first
hours is assigned to the mechanical loss of poorly attached
CoHCF crystallites, which becomes less relevant as the experiment proceeds. The current density reaches c.a. 50 % and 20 %
of its initial value after c.a. 4 and 50 hours, respectively.
However, this phenomenon presumably takes place only at the
very first use of the electrode and does not lead to the full
depletion of the active sites as it was confirmed by the postreaction characterization shown above. AAO/FTO/PB shows to
be stable with a decay of the current density of 0.17 mA cm-2
for the last 40 hours. A steady-state current density of
1.85 mA cm-2 is obtained after 100 hours polarization at 2.2 V
vs, RHE without any further significant sign of deterioration.
This denotes the long-term robustness of the proposed
electrode, which is not comparable with the robustness
reported in most water oxidation studies in half-cell configuration, typically evaluated for less than 24 h. One can find in
Supplementary information a video (Video S1) where oxygen
bubbles evolving from the electrode surface during the CA at
2.2 V vs. RHE can be observed.
The stability of the AAO/FTO/PB electrode has also been
confirmed at 2.4 V vs. RHE by using a new AAO/FTO/PB
electrode prepared in a similar way than the previous one. A
stable current density of 2.6 mA cm-2 is obtained at this voltage
after 100 hours (See Figure S2 in Supplementary Information). A
faradaic efficiency towards oxygen evolution reaction closed to
100 % has been confirmed by in-situ measuring, during the
latter CA, the increase of oxygen concentration in the anode
gas headspace using a FOXY probe. The similarity between the
O2 production experimentally measured during a given polarization time and the theoretical (faradaic) O2 production
calculated from the electric charge transferred during the same
period can be clearly observed in Figure S3 in Supplementary
Information.
In order to prove the superior properties of the AAO/FTO/
PB electrode with respect to an analogous catalyst prepared on
a flat surface support, i. e. instead of AAO, additional electrodes
have been prepared by FTO deposition on glass substrates and
subsequent Prussian blue catalyst deposition by following a
similar procedure than that carried out on AAO supports. ESEM
images of both glass- and AAO-supported FTO layers are
observed in Figure S4. A smoother FTO surface is apparently
obtained on flat glass compared to the FTO layer grown on
AAO where a higher roughness can be appreciated because of
a greater presence of defects on the latter substrate. Figure S5
shows the EDX and XRD characterization of the Glass/FTO/PB
electrode and the corresponding blank electrode (i. e., in
absence of PB). A much lower presence of CoHCF on the Glass/

FTO surface can be deduced with respect to the EDX spectra
and XRD diffractograms obtained with the AAO-supported
electrode (i. e., Figure 6), which can be attributed to a stronger
attachment of the catalyst nanoparticles on the rough surface
of the latter electrode. The same electrochemical tests than
those shown in Figures 8–10 have been performed on both
Glass/FTO/PB and Glass/FTO (See Figures S6–8 in Supplementary information). From the linear sweep voltammetry measurement (Figure S6) much lower current density values can be
observed in the same voltage scan than those values previously
shown in Figure 8 with AAO-supported catalyst, as expected
due to the lower roughness of the FTO layer deposited on the
glass support. Higher Tafel slope is also obtained with Glass/
FTO/PB (Figure S7) than with AAO/FTO/PB (Figure 9). Finally, in
the chronoamperometry performed at 2.2 V vs. RHE for 100 h
with the glass-supported electrode (Figure S8), a continuous
decrease of the current density has been observed during the
whole polarization, reaching a final current density of i. e.,
0.78 mA cm-2, in contrast to the stable current density obtained
with the AAO-supported electrode, i. e., 1.85 mA cm-2 (Figure 10). Thus, the use of anodic aluminum oxide as catalyst
support is confirmed to provide not only higher electrocatalytic
activity but also higher electrode stability. We attribute this
effect to the higher FTO layer roughness obtained on AAO
substrate and the stronger attachment of the PB nanoparticles
on this surface.
If we consider the electrocatalytic performances reported in
benchmarking studies of water oxidation catalysts,[46–47] in terms
of overpotential required to obtained a given current density
after 2 or 24 hours, it is clear that the results obtained with the
studied AAO/FTO/PB electrode at pH 7 are still not comparable
with those obtained with either Ru and Ir catalysts at pH 0 or
electrodes based on Co, Ni and Fe, among others, at pH 14.
However, from the obtained results it can be stated that the
electrocatalytic activity of the AAO/FTO/PB catalyst for water
oxidation is higher than that previously reported with FTO/PB
electrodes[28–29] and other earth-abundant materials-based
electrodes.[48–49] Thus, the feasibility of FTO-covered AAO as a
conductive and efficient catalyst support for electrocatalytic
processes is demonstrated herein and, particularly, for water
electrolysis at neutral pH and ambient conditions. It should also
be noted that the proposed configuration is prone to further
optimization, by tuning the porous structure of the AAO and by
improving the deposition method for both the FTO crystallites
and the catalyst nanoparticles.

3. Conclusions
An efficient and simple route to achieve a significant electrical
conductivity on an alumina surface (AAO) was established by
FTO deposition via spray pyrolysis. From the results of X-ray
diffraction, I-V curve and ESEM images we conclude that the
as-deposited FTO crystallite size increases with deposition time,
while the film electrical resistivity decreases down to a basal
minimum. We decorated an optimized FTO/AAO composite
with Prussian blue (PB)-type cobalt hexacyano-ferrate nano-

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57

particles, a promising earth-abundant metal-based electrocatalyst for water oxidation, exhibiting excellent performance in
terms of electrochemical activity and stability during the
oxygen evolution reaction. Our results open interesting possibilities for the development of electrode supports from intrinsic
insulators, possessing very high surface areas, such as AAO.
Taking advantage of the latter, surface modification with a
conducting layer allows their transformation into functional
electrodes. In neutral media, our AAO/FTO/PB composite
electrodes promote electrochemical water splitting at higher
current densities that any other non-noble metal-based electrodes, with long term stability (> 100 h). In this proof of concept,
we also observed a synergic relation between the porous
alumina and the conducting phase, with enhanced FTO
conductivity due to the presence of induced defects in
conducting component.

Experimental Section
Synthesis of the Electrodes
Fluorine-doped tin oxide was deposited on AAO substrates. The
precursor solution was synthesized from Sn(IV) salts, using
SnCl4·5H2O. A 0.23 M solution was prepared in a mixture of polar
solvents: CH3CH2OH 99.9 % and deionized H2O in a ratio of 9 : 1,
respectively. 0.27 g of NH4F were used as the dopant source, to
obtain a concentration of 5 wt.% of F with respect to the amount of
Sn present in the stock solution, i. e. resulting in a F/Sn = 0.05 ratio.
Subsequently the ethanolic solution was heated at 80 °C for 60 min
to hydrolyze the Sn + 4 ion and, in this way, obtain the Sn(OH)
x
solution. This precursor solution was deposited onto AAO membranes using a home-made spray pyrolysis system. The configuration scheme of a similar system has been previously reported.[50]
AAO substrates with pores around 220 nm of diameter and 60 μm
of thickness (Sigma-Aldrich) were used. The deposition temperature, controlled within � 1°C, was set at 440 °C and the deposition
times were in the range from 10 to 120 min including deposition
on both sides of the AAO substrate (i. e. from 5 to 60 min on each
side). FTO was also deposited on glass substrates at the same
deposition times. The latter samples were used to obtain the FTO
deposition rate by estimating the optical thickness in each case
from the transmittance spectrum, as reported elsewhere.[35] FTO on
glass was also deposited as reference support for the water
oxidation catalyst, in order to compare the electrocatalytic performance of AAO-supported electrodes and those supported on a flat
surface.
Prussian blue catalyst was obtained via the dip coating technique.
Two solutions were prepared for the depositions. Solution A: 0.01 M
K3[Fe(CN)6], and solution B: 0.01 M Co(NO3)2·6H2O. Five deposition
cycles were performed, where AAO/FTO and Glass/FTO samples
were dipped, first, in solution A and, second, in solution B for
15 minutes in each case, followed by air drying overnight. The
deposition process was carried out under magnetic stirring and
only an area of 1 cm2 from each electrode was dipped. All chemicals
used in this study were of analytical grade and provided by SigmaAldrich.

the Cu Kα line (λ = 1.541 Å). ESEM images and EDX spectra were
acquired in a QUANTA600 equipment from FI company under high
vacuum conditions. The electron microscope was operated at
20 kV. This physicochemical characterization was carried out both
before and after the electrocatalytic tests. I–V measurements were
performed in a coplanar configuration using a Keithley 2400 digital
picoammeter/voltage source from Keithley Instruments, Inc.. The
voltage was applied between the silver epoxy contacts spaced
1 cm on each surface and the current was collected using the same
equipment. So, to obtain the sheet (in-plane) electrical resistance of
the different FTO layers between these points, the slope of the
resultant I–V curve was used to obtain the surface resistance value
corresponding to each coating condition. Then the respective
resistivities were estimated by considering the FTO optical thicknesses, as was also accomplished in previous works.[35]

Electrocatalytic Tests
The electrochemical experiments were performed at room temperature and atmospheric pressure in half-cell configuration (See
Figure S9 in Supplementary Information) by using a Biologic SP-150
potentiostat and a buffer electrolyte prepared in 50 mM KPi (pH 7)
containing 1 M KNO3. A H-cell was used where the anode and
cathode compartments were separated by a porous glass frit and
all the experiments were carried out under magnetic stirring at c.a.
1000 rpm in order to remove the O2 and H2 bubbles from the
electrodes (see Video S1 in Supplementary Information). Linear
Sweep Voltammetry (LSV) and Chronoamperometry (CA) measurements were performed by employing Ag/AgCl (3.5 M KCl) from ALS
as reference electrode, Pt mesh as counter electrode (cathode) and
AAO/FTO/PB or Glass/FTO/PB as working electrode (anode). Blank
measurements were also carried out with AAO/FTO and Glass/FTO
samples. Before the electrochemical measurements the electrical
contacts were made by tin welding and silver conductive epoxi
from RS COMPONENTS LTD. on the Glass- and AAO-supported
electrodes, respectively, in electrode areas not covered by Prussian
blue. In all cases the electrical contacts were finally coated by epoxy
adhesive from RS COMPONENTS LTD and were avoided to dip into
the electrolyte during the electrocatalytic tests.
All potentials measured in this work were iR corrected by the
current interruption technique and converted to the reversible
hydrogen electrode (RHE) scale through equation 1.

EðRHEÞ ¼ EðAg=AgClÞ þ 0:059 pH þ E0 ðAg=AgClÞ

(1)

Where E (Ag/AgCl) is the measured working potential and E0 (Ag/AgCl) is
0.205 V at 25 °C. Then overpotential (η) was calculated from
equation 2.

h ¼ EðRHEÞ- E 0cell

(2)

Where E0cell is the standard potential of the water electrolysis cell
which is 1.229 V at 25 °C. All current density values have been
normalized per electrode geometrical area. The experimental
procedure carried out for the in-situ oxygen evolution measurement is described in Supplementary information.

Acknowledgements
Characterization of the Electrodes
XRD patterns were obtained from the different samples by using a
Siemens EM-10110BU model D5000 diffractometer, operating with

We would like to acknowledge financial support from the Spanish
Ministerio de Economía y Competitividad (MINECO) through
project CTQ2015-71287-Rand the Severo Ochoa Excellence Accred-

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57

itation 2014-2018 SEV-2013-0319; and the Generalitat de Catalunya (2017-SGR-1406 and the CERCA Programme). FAGP thanks
the Marie Curie COFUND Action from the European Commission
for co-financing his postdoctoral fellowship.

[25]
[26]
[27]
[28]
[29]

Conflict of Interest

[30]

The authors declare no conflict of interest.
[31]

Keywords: anodic aluminium oxide · fluorine-doped tin oxide ·
Prussian Blue · water electrolysis · water electrooxidation

[1] W. Lee, S. J. Park, Chem. Rev. 2014, 114, 7487–7556.
[2] A. M. Md Jani, D. Losic, N. H. Voelcker, Prog. Mater. Sci. 2013, 58, 636–
704.
[3] Z. A. Hamid, M. T. A. El-Khair, M. H. Gomaa, F. A. Morsy, N. A. A. Khalifa,
Int. J. Nanopart. 2014, 7, 231–250.
[4] H. Zhao, M. Zhou, L. Wen, Y. Lei, Nano Energy 2015, 13, 790–813.
[5] H. Dang, V. P. Singh, S. Guduru, S. Rajaputra, Z. D. Chen, Nano Res. 2015,
8, 3186–3196.
[6] H. Zhang, J. H. Kim, J. S. Lee, Adv. Funct. Mater. 2017, 27, 1702428.
[7] A. Brzózka, A. Jeleń, A. M. Brudzisz, M. M. Marzec, G. D. Sulka, Electrochim. Acta 2017, 225, 574–583.
[8] H. E. Cheng, S. Y. Lin, D. C. Tian, J. Electrochem. Soc. 2014, 161, E202–
E206.
[9] S. Garbarino, A. Ponrouch, E. Bertin, D. Guay, Mater. Res. Soc. Symp. Proc.
2011, 1311, 7–12.
[10] W. Yan, D. Wang, G. G. Botte, J. Appl. Electrochem. 2015, 45, 1217–1222.
[11] T. Y. Shin, S. H. Yoo, S. Park, Chem. Mater. 2008, 20, 5682–5686.
[12] H. Wang, C. Xu, F. Cheng, M. Zhang, S. Wang, S. P. Jiang, Electrochem.
Commun. 2008, 10, 1575–1578.
[13] J. Cui, Y. Wu, Y. Wang, H. Zheng, G. Xu, X. Zhang, J. Nanosci.
Nanotechnol. 2013, 13, 1149–1152.
[14] S. Altuntas, F. Buyukserin, Appl. Surf. Sci. 2014, 318, 290–296.
[15] E. Wierzbicka, G. D. Sulka, Sens. Actuators B 2016, 222, 270–279.
[16] M. Schalenbach, A. R. Zeradjanin, O. Kasian, S. Cherevko, K. J. J.
Mayrhofer, Int. J. Electrochem. Sci. 2018, 13, 1173–1226.
[17] S. Y. Tee, K. Y. Win, W. S. Teo, L. D. Koh, S. Liu, C. P. Teng, M. Y. Han, Adv.
Sci. 2017, 4, 1600337.
[18] J. Qi, W. Zhang, R. Cao, ChemCatChem 2018, 10, 1206–1220.
[19] S. Park, Y. Shao, J. Liu, Y. Wang, Energy Environ. Sci. 2012, 5, 9331–9344.
[20] E. Özer, C. Spöri, T. Reier, P. Strasser, ChemCatChem 2017, 9, 597–603.
[21] X. Liang, R. Dong, D. Li, X. Bu, F. Li, L. Shu, R. Wei, J. C. Ho, ChemCatChem
2018, 10, 4555–4561.
[22] Y. Yi, G. Weinberg, M. Prenzel, M. Greiner, S. Heumann, S. Becker, R.
Schlögl, Catal. Today 2017, 295, 32–40.
[23] S. Schlicht, S. Haschke, V. Mikhailovskii, A. Manshina, J. Bachmann,
ChemElectroChem 2018, 5, 1259–1264.

[32]

[33]
[34]
[35]
[36]
[37]
[38]
[39]
[40]
[41]
[42]
[43]

[44]
[45]
[46]
[47]
[48]
[49]
[50]

[24] N. Linck, A. Peek, B. J. Hinds, ACS Appl. Mater. Interfaces 2017, 9, 30964–
30968.
J. R. Galán-Mascarós, ChemElectroChem 2014, 2, 37–50.
M. I. Jamesh, X. Sun, J. Power Sources 2018, 400, 31–68.
M. Aksoy, S. V. K. Nune, F. Karadas, Inorg. Chem. 2016, 55, 4301–4307.
E. P. Alsaç, E. Ülker, S. V. K. Nune, Y. Dede, F. Karadas, Chem. Eur. J. 2018,
24, 4856–4863.
S. Pintado, S. Goberna-Ferrón, E. C. Escudero-Adán, J. R. Galán-Mascarós,
J. Am. Chem. Soc. 2013, 135, 13270–13273.
F. S. Hegner, I. Herraiz-Cardona, D. Cardenas-Morcoso, N. López, J. R.
Galán-Mascarós, S. Gimenez, ACS Appl. Mater. Interfaces 2017, 9, 37671–
37681.
B. Rodríguez-García, Á. Reyes-Carmona, I. Jiménez-Morales, M. BlascoAhicart, S. Cavaliere, M. Dupont, D. Jones, J. Rozière, J. R. GalánMascarós, F. Jaouen, Sustainable Energy Fuels 2018, 2, 589–597.
L. Han, P. Tang, Á. Reyes-Carmona, B. Rodríguez-García, M. Torréns, J. R.
Morante, J. Arbiol, J. R. Galan-Mascaros, J. Am. Chem. Soc. 2016, 138,
16037–16045.
D. Mohanta, M. Ahmaruzzaman, RSC Adv. 2016, 6, 110996–111015.
R. A. Afre, N. Sharma, M. Sharon, Rev. Adv. Mater. Sci. 2018, 53, 79–89.
F. A. Garcés, N. Budini, R. R. Koropecki, R. D. Arce, Thin Solid Films 2013,
531, 172–178.
E. Ching-Prado, A. Watson, H. Miranda, J. Mater. Sci. Mater. Electron.
2018, 29, 15299–15306.
K. H. Kim, B. R. Koo, H. J. Ahn, Ceram. Int. 2018, 44, 9408–9413.
Y. Gao, Y. Lin, J. Chen, Q. Lin, Y. Wu, W. Su, W. Wang, Z. Fan, Nanoscale
2016, 8, 13280–13287.
A. E. Hassanien, H. M. Hashem, G. Kamel, S. Soltan, A. M. Moustafa, M.
Hammam, A. A. Ramadan, Int. J. Thin Film. Sci. Tec. 2016, 5, 55–65.
H. M. Yates, P. Evans, D. W. Sheel, Z. Remeš, M. Vanecek, Ü. Dagkaldiran,
A. Gordijn, F. Finger, Int. J. Nanotechnol. 2009, 6, 816–827.
A. Bleuzen, C. Lomenech, V. Escax, F. Villain, F. Varret, C. Cartier Dit
Moulin, M. Verdaguer, J. Am. Chem. Soc. 2000, 122, 6648–6652.
M. Afzaal, H. M. Yates, A. Walter, S. Nicolay, C. Ballif, J. Mater. Sci. 2017, 5,
4946–4950.
F. Z. Dahou, L. Cattin, J. Garnier, J. Ouerfelli, M. Morsli, G. Louarn, A.
Bouteville, A. Khellil, J. C. Bernède, Thin Solid Films 2010, 518, 6117–
6122.
J. D. Benck, B. A. Pinaud, Y. Gorlin, T. F. Jaramillo, PLoS One 2014, 9,
e107942.
S. Geiger, O. Kasian, A. M. Mingers, K. J. J. Mayrhofer, S. Cherevko, Sci.
Rep. 2017, 7, 4595.
C. C. L. McCrory, S. Jung, J. C. Peters, T. F. Jaramillo, J. Am. Chem. Soc.
2013, 135, 16977–16987.
C. C. L. McCrory, S. Jung, I. M. Ferrer, S. M. Chatman, J. C. Peters, T. F.
Jaramillo, J. Am. Chem. Soc. 2015, 137, 4347–4357.
M. Martin-Sabi, J. Soriano-López, R. S. Winter, J. J. Chen, L. Vilà-Nadal,
D. L. Long, J. R. Galán-Mascarós, L. Cronin, Nat. Catal. 2018, 1, 208–213.
S. Haschke, Y. Wu, M. Bashouti, S. Christiansen, J. Bachmann,
ChemCatChem 2015, 7, 2455–2459.
B. Gottlieb, R. Koropecki, R. Arce, R. Crisalle, J. Ferron, Thin Solid Films
1991, 199, 13–21.