RESEARCH ARTICLE
www.advancedscience.com
Efficient Decoupled Electrolytic Water Splitting in Acid
through Pseudocapacitive TiO2
Mairis Iesalnieks, Ma¯rtin¸š Vanags, Linda Laima Alsin¸a, Raivis Egl¯ıtis, L¯ıga Gr¯ınberga,
Peter C. Sherrell,* and Andris Šutka*
Water electrolysis remains a key component in the societal transition to green
energy. Membrane electrolyzers are the state-of-the-art technology for water
electrolysis, relying on 80 °C operation in highly alkaline electrolytes, which is
undesirable for many of the myriad end-use cases for electrolytic water
splitting. Herein, an alternative water electrolysis process, decoupled
electrolysis, is described which performed in mild acidic conditions with
excellent efficiencies. Decoupled electrolysis sequentially performs the oxygen
evolution reaction (OER) and the hydrogen evolution reaction (HER), at the
same catalyst. Here, H+ ions generated from the OER are stored through
pseudocapacitive (redox) charge storage, and released to drive the HER. Here,
decoupled electrolysis is demonstrated using cheap, abundant, TiO2 for the
first time. To achieve decoupled acid electrolysis, ultra-small anatase TiO2
particles (4.5 nm diameter) are prepared. These ultra-small TiO2 particles
supported on a carbon felt electrode show a highly electrochemical surface
area with a capacitance of 375 F g−1. When these electrodes are tested for
decoupled water splitting an overall energy efficiency of 52.4% is observed,
with excellent stability over 3000 cycles of testing. This technology can
provide a viable alternative to membrane electrolyzers—eliminating the need
for highly alkaline electrolytes and elevated temperatures.
M. Iesalnieks, M. Vanags, L. L. Alsin¸a, R. Egl¯ıtis, A. Šutka
Institute of Materials and Surface Engineering
Faculty of Natural Sciences and Technology
Riga Technical University
P. Valdena Street 3/7, Riga LV-1048, Latvia
E-mail: andris.sutka@rtu.lv
L.Gr¯ınberga
Institute of Solid State Physics
University of Latvia
Riga LV-1063, Latvia
P. C. Sherrell
AppliedChemistry&Environmental Science
School of Science
RMITUniversity
124 LaTrobeSt,Melbourne 3000, Australia
E-mail: peter.sherrell@rmit.edu.au
The ORCID identification number(s) for the author(s) of this article
can be found under https://doi.org/10.1002/advs.202401261
© 2024 The Authors. Advanced Science published by Wiley-VCH GmbH.
This is an open access article under the terms of the Creative Commons
Attribution License, which permits use, distribution and reproduction in
any medium, provided the original work is properly cited.
DOI: 10.1002/advs.202401261
1. Introduction
Hydrogen is considered a versatile and
clean energy carrier as it can be pro-
duced by water electrolysis, electrochemi-
cally splitting water molecules into hydro-
gen and oxygen.[1] Green hydrogen is pro-
duced when electrolysis is performed using
only renewable energy such as solar or wind
power[2] and can help with grid equalizing
in place of batteries.[3] The produced hydro-
gen can then be stored prior to use as a
clean fuel, reacting with oxygen to produce
water, with zero greenhouse gas emissions,
being touted for many applications, includ-
ing fuel cells in transport,[4] for blending
with natural gas for combustion,[5] and
for chemical production via the Fischer–
Tropsch process.[6]
Currently, commercial water electroly-
sis is performed predominantly in mem-
brane electrolyzers. However, these mem-
brane electrolyzers have several drawbacks,
including high material and catalyst cost,
high assembly cost due to device complex-
ity, high maintenance costs, low tolerance
to impurities in the electrolyte,[7,8] and adverse operating condi-
tions (high temperature, highly alkaline electrolytes), which in
turn lead to membrane fouling.[9] The membrane electrolysis
concept also raises operational safety issues due to H2/O2 cross-
over across the membrane/separator.
Recently, the decoupled water electrolysis concept has
emerged as an alternative to membrane electrolysis. In de-
coupled electrolysis the oxygen evolution reaction (OER) and
hydrogen evolution reaction (HER) are performed sequentially
on the same catalyst, thus avoiding the use of membrane,
enabling non-alkaline electrolytes, a wide window of operating
conditions, and lowers the requirements related to auxiliary
devices for maintenance.[10] These advantages make decoupled
water electrolysis a promising alternative to explore for achieving
green hydrogen production.
In decoupled electrolysis, two main strategies have been ex-
plored, by harnessing electron-coupled proton buffers[11–14]; or
by utilizing pseudocapacitive electrodes to accumulate OH− or
H+ ions while they are not being consumed in the catalytic
reactions.[9,15,16] In pseudocapacitive systems, the OH− or H+
ions are stored while the OER is occurring, and released while
HER is occurring.
Adv. Sci. 2024, 11, 2401261 © 2024 The Authors. Advanced Science published by Wiley-VCH GmbH2401261 (1 of 10)
www.advancedsciencenews.com www.advancedscience.com
Figure 1. Concept of decoupled water splitting using a TiO2 auxiliary electrode; a) visual representation of the electrochemical cell setup; and b) reaction
diagram showing cyclical HER and OER reactions.
Literature reports of these systems have focused on two main
classes of materials, Ni(OH)2-based pseudocapacitors have been
exploited for alkaline electrolytes, whileWO3-based pseudocapac-
itors have been exploited for acid-decoupled electrolysis.[17–19]
In the case of Ni(OH)2, in step (1), the H2 is released follow-
ing reaction 4H2O + 4e−→ 4OH− + 2H2, with OH− intercalat-
ing into the Ni(OH)2, which in turn is transformed to NiOOH
by the reaction Ni(OH)2 + OH− → NiOOH + H2O + e−9. The
NiOOH is transformed back to Ni(OH)2 in step (2), and oxygen is
released. In the case of WO3, oxygen is released in step (1) along-
side H+ intercalation into WO3 via 2H2O → 4H
+ + 4e− + O2
and xH+ + xe− + WO3 → HxWO3, respectively. In step (2),
H+ is released from the WO3 pseudocapacitive electrode and
H2 is released on the working electrode by the reactions
HxWO3 → xH
+ + xe− +WO3 and 2H+ + 2e− →H2, respectively.
However, while these approaches are promising, both Ni and W
present challenges moving forward, with Ni(OH)2 exhibiting sta-
bility issues[15] along with environmental and health risks, andW
being considered by the EuropeanUnion as a criticalmetal with a
high supply risk.[20] Recently, carbonaceous materials have been
reported for such decoupled electrochemical systems, either di-
rectly for water splitting,[7] or using the same fundamental con-
cept as a hydrogen battery.[21] These recent works highlight the
opportunity advantages in the use of cheap, abundant, electrode
materials for decoupled electrolysis.
TiO2 is one of the most abundant oxides globally and is an ex-
cellent pseudocapacitive material that has not been studied for
decoupled electrolysis. While small TiO2 particles have demon-
strated capability for Li+ [22,23] and Na+ intercalation,[24] H+ inter-
calation into TiO2 required for decoupled electrolysis is scarcely
reported.[25] Kim et al.[26] demonstrated H+ intercalation in high
surface area mesoporous amorphous TiO2 electrode films via a
proton-coupled electron transfer reaction, equivalent to the Li+-
coupled electron transfer reaction.[27] In other work, TiO2 nan-
otube layers showedH+ intercalation/de-intercalation ability dur-
ing cathodic/anodic cycling with a high storage capacity.[28] These
preliminary suggest the ability of TiO2 to support decoupled wa-
ter electrolysis.
Herein, we prepare ultra-small (<5 nm diameter) TiO2
nanoparticles in a conductive carbon blackmatrix and probe their
performance in decoupled water electrolysis.[29] The ultra-small
nanoparticles demonstrate excellent stability over 3000 cycles and
an overall electrolysis efficiency of 52.4%, thus opening new op-
portunities for developing ecofriendly hydrogen generation by ex-
ploiting abundant materials.
2. Results and Discussion
2.1. Concept of Decoupled Water Electrolysis in Acidic Media
The decoupled electrolysis principle using pseudocapacitive TiO2
in an acidic 0.5 m H2SO4 solution is demonstrated in Figure
1a. A platinum electrode was used as a model system to cycli-
cally catalyze the 1) OER or 2) HER. To control which reac-
tion occurs on the Pt catalyst, and how ions and charge are
stored at the TiO2 auxiliary electrode, the polarity of the ap-
plied voltage was changed stepwise. In the first water-splitting
cycle, step (1), TiO2 is protonated to titanium oxyhydroxide
(TiO2 + e− + H+ → Ti(O)(OH)) in acidic media[30] following the
Pourbaix diagram.[31] Step (1) is energetically costly as the OER
(2H2O→ O2 + 4e− + 4H+), requiring four electrons to proceed.
When the cell polarity is reversed, in step (2), the H+ ions stored
at the TiO2 electrode in step (1) are released to enable the HER
on the Pt catalyst. Step (2) then continues until Ti(O)(OH) is fully
converted back to TiO2, where the cycle is repeated (Figure 1b).
As gases are generated from ions released from the TiO2 elec-
trode, and HER occurs in acidic media, hydrogen generation at
the second step is theoretically highly energy efficient/beneficial,
as described below.
2.2. TiO2 Synthesis and Characterization
To achieve the concept described above, high surface area TiO2
is required. TiO2 was produced sol–gel synthesis (see Experi-
mental Section for full details), producing ultra-small anatase
nanoparticles. The anatase structure was confirmed by Raman
spectroscopy (Figure 2a) and X-ray diffraction (XRD) (JCPDS
21–1272) (Figure 2b). The Raman spectra in Figure 2a show
peaks at 148.17 cm−1 (Eg), 198.34 cm
−1 (Eg), 398.13 cm
−1 (B1g),
515.36 cm−1 (A1g), and 638.85 cm
−1 (Eg) corresponding to TiO2
anatase vibrational modes.[32] A slight Raman band shift can be
observed in Raman spectroscopy due to the small size of synthe-
sized nanoparticles compared to the literature.[33] The crystallites
Adv. Sci. 2024, 11, 2401261 © 2024 The Authors. Advanced Science published by Wiley-VCH GmbH2401261 (2 of 10)
 21983844, 2024, 28, D
ow
nloaded from
 https://onlinelibrary.w
iley.com
/doi/10.1002/advs.202401261 by Cochrane Latvia, W
iley O
nline Library on [07/01/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on W
iley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
www.advancedsciencenews.com www.advancedscience.com
Figure 2. Characterization of synthesized TiO2 nanoparticles after surfactant stripping; a) Raman spectroscopy; b) XRD diffractogram with JCPDS 21–
1272 standard; c) TEM image of synthesized particles; d) Histogram of particle size distribution obtained from analysis of 120 particles via TEM; and
high-resolution XPS spectra for; e) titanium 2p; and f) oxygen 1s.
size was calculated to be 4.54 nm from the average of all XRD
peaks, via the Scherrers’ equation:[34]
D = k𝜆
B cos 𝜃
(1)
where D is the crystallite size (nm); k is the particle form factor
(0.9); 𝜆 is the wavelength of the X-ray source (0.154056 nm); B is
the peak full width at half maximum height; and 𝜃 is the Braggs
angle.
Transmission electron microscopy (TEM) analysis (Figure 2c)
shows individual TiO2 nanoparticles, with an average size of
3.90 ± 0.17 nm, aligned with the calculated size from XRD.
X-ray photoelectron spectroscopy (XPS) of the Ti 2p and O 1s
regions are shown in Figure 2e,f respectively. Binding energies
of the Ti 2p3/2 peak at 458.9 EV indicate the presence of a Ti
4+ ox-
idation state without a lower oxidation state- Ti3+. Peak splitting
between Ti 2p3/2 and Ti 2p1/2 was determined to be 5.7 eV, in-
dicating the presence of anatase.[35] Oxide lattice oxygen formed
the majority of O 1s spectra, with ≈26 (±1%) of ─OH surface hy-
droxyl component at a binding energy of 531.1 eV, corresponding
to the first couple of monolayers of surface hydroxyl.[36]
2.3. Electrochemical Properties of Auxiliary TiO2 Electrodes
To identify the suitability of the anatase TiO2 particles for the
acidic, decoupled electrochemical water splitting reaction cyclic
voltammetry (CV) was performed.
In Figure 3a, the CV of a commercial “Aeroxide P25” TiO2
electrode is shown (Figure S3, Supporting Information), these
particles have a diameter of ≈25 nm. Between an applied volt-
age of −0.3 and 0.4 V versus the saturated calomel electrode
(SCE), only electrochemical double-layer capacitance is observed.
Adv. Sci. 2024, 11, 2401261 © 2024 The Authors. Advanced Science published by Wiley-VCH GmbH2401261 (3 of 10)
 21983844, 2024, 28, D
ow
nloaded from
 https://onlinelibrary.w
iley.com
/doi/10.1002/advs.202401261 by Cochrane Latvia, W
iley O
nline Library on [07/01/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on W
iley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
www.advancedsciencenews.com www.advancedscience.com
Figure 3. Cyclic voltammetry (CV) curves for a) TiO2 P25, b) TiO2 with surfactant, and c) TiO2 without surfactant nanoparticle electrodes. Curves were
taken in 0.5 m H2SO4 electrolyte in a three-electrode cell with a scan rate of 5 mV s
−1.
However, as the voltage drops below −0.3 V versus SCE, the ca-
thodic current increases rapidly representing electrolyte break-
down, in this case, the HER. Hydrogen gas bubbles were visually
observed to be released below this voltage, so the voltage in the
cycle was limited to −0.5 V.
In comparison, the CV curves of the as synthesized ultra-small
TiO2 nanoparticles (Figure 3b,c) differ markedly from the CV
curve of the TiO2 P25 sample. Here, two configurations of ultra-
small TiO2 nanoparticle electrodes were studied; as produced
(Figure 3b), and after stripping of the surfactant using hexane
to expose the pristine TiO2 surface (Figure 3c). The as-produced
particles remain partially coated in the surfactant used for syn-
thesis, 4-dodecylbenzene sulphonic acid (4-DDBSA). For the as-
produced TiO2 nanoparticles (with surfactant), as the potential
decreased below −0.25 V versus SCE, a pronounced cathodic
current appears, which forms an indistinct reduction peak at
≈−0.38 V versus SCE. A secondary reduction peak is observed
−0.58 V versus SCE, with hydrogen gas evolution being observed
at potentials below −0.7 V versus SCE. As the potential is in-
creased, an oxidation peak appears at −0.5 V, followed by another
oxidation peak at −0.13 V, followed by electrochemical double
layer formation.
In the CV curve of the stripped surfactant-free TiO2 NPs
(Figure 3c), the reduction and oxidation peaks clearly appear at
−0.4 V versus SCE and −0.13 V versus SCE, respectively. This
redox process corresponds to the following reaction:[37]
TiO2 + e− +H3O+ ↔ Ti (O)OH +H2O (2)
To understand the different CV shapes for P25, and ultra-
small TiO2 particles with and without surfactant, the H
+ inter-
calation process must be considered. H+ intercalation can only
occur at the surface of TiO2 particles,
[38] and therefore the sur-
face area:volume ratio of these systems is a key parameter. This
surface area:volume ratio inversely scales with particle size, and
thus, in the case of the large P25 TiO2 particlesminimal H
+ inter-
calation can occur. In contrast, the ultra-small TiO2 particles have
a dramatically increased surface area to volume ratio, enabling
significant H+ intercalation to occur for the same mass of parti-
cles. Further, the stripping of surfactant exposes the full surface
area of the ultra-small TiO2 samples to the electrolyte, leading to
optimize H+ intercalation and thus a maximum for the observed
redox peaks. Thus, the stripped ultra-small TiO2 nanoparticles ex-
hibit the strongest redox features, showing that the reaction (1)
dominates in the particles (Figure 3c).
Electrochemical impedance spectroscopy (EIS) studies were
performed at open circuit potential in the same cell used in CV
measurements to understand the electrolyte–electrode interface
(Figure 4). Two incomplete semicircles characterize plots for all
three samples. A semicircle of a small diameter appears in the
high-frequency region, while a semicircle of a large diameter ap-
pears in themid and low-frequency region. For eachNyquist plot,
the equivalent circuit was simulated (Figure 4 inset).
The circuit consists of two parallel circuits of resistance (R)
and constant phase elements (CPEs), namely R2, CPE1, and R3
CPE2. The scheme also has a series resistance R1, which char-
acterizes the resistance of the wires, the conductive substrate
and the electrolyte.[39] The first parallel connection, consisting
of CPE1 and R2, characterizes the solvent molecules’ geometric
capacitance and orientational polarization resistance. This con-
nection purely characterizes the surface boundary layer between
electrode and electrolyte with no Faradaic response. The second
parallel connection, R3 and CPE2, characterizes the double-layer
capacitance on the boundary surface of the electrode-electrolyte
and the charge transfer resistance.[40] This charge transfer resis-
tance is related to the electron transfer resistance of the reaction
Figure 4. Nyquist plots in complex plane for all three TiO2 samples. (Inset:
fitted equivalent circuit model).
Adv. Sci. 2024, 11, 2401261 © 2024 The Authors. Advanced Science published by Wiley-VCH GmbH2401261 (4 of 10)
 21983844, 2024, 28, D
ow
nloaded from
 https://onlinelibrary.w
iley.com
/doi/10.1002/advs.202401261 by Cochrane Latvia, W
iley O
nline Library on [07/01/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on W
iley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
www.advancedsciencenews.com www.advancedscience.com
Table 1. Equivalent circuit parameter values from Nyquist plots.
Parameter TiO2 P25 TiO2 with surfactant TiO2 without surfactant
Value Error [%] Value Error [%] Value Error [%]
R1 Ω 3.1 ±0.5 0.5 ±1.5 2.0 ±0.7
R2 Ω 1.5 ±3.1 2.1 ±3.4 1.1 ±1.1
CPE1 Y1, Ω∙sN1 0.327 ±7.4 0.946 ±9.3 9.56 ±8.7
N1 0.58 ±6.2 0.35 ±7.4 0.40 ±2.3
R3 Ω 82.7 ±6.5 35.8 ±7.2 27.4 ±5.7
CPE2 Y2, Ω∙sN2 26.7 ±1.2 26.2 ±2.1 229 ±0.6
N2 0.82 ±0.6 0.83 ±0.8 0.80 ±0.5
(2). By analyzing the values of the simulated parameters summa-
rized in Table 1, we can observe that the value of R1 is approx-
imately similar for all samples since the wires, base, and elec-
trolyte were the same in all three cases. The R2 values are also
similar in the border of error for all three samples since the sol-
vent was water in all cases.
Noticeable differences between the samples appear in the
CPE1 parameters Y1 and N1. According to the obtained results,
the Y1 value is 0.327 for the TiO2 P25 sample, 0.946 for the ultra-
small TiO2 and 9.56 for the stripped ultra-small TiO2. Y1 almost
inherits capacitance, but if in an ideal capacitor, the phase shift is
𝜋
2
, then the phase shift of the CPE isN ⋅ 𝜋
2
, where N varies from 0
to 1.[7] Therefore, Y values will be used for capacity characteriza-
tion in the following because the N values are sufficiently similar
to the corresponding CPE. The small capacity of the TiO2 P25
sample can be explained by the much larger particle size (25 nm)
compared to the ultra-small TiO2 particles (≈4 nm). Therefore,
the surface area in contact with the electrolyte for such small par-
ticles is much larger, and the surface capacitance also increases.
There is also a significant difference in the surface characteristic
capacitance parameter between the synthesized ultra-small TiO2
before and after surfactant stripping. This difference arises as the
surfactant prevents the electrolyte from interacting with the ac-
tive TiO2 surface, leading to decreased capacitance and the elec-
trode | electrolyte interface. The R3 parameter in Table 1 is the
smallest for synthesized ultra-small TiO2 stripped from surfac-
tant due to the lower charge transfer resistance. The ion interca-
lation also occurs on the surface of large-size TiO2 nanoparticles.
The charge transfer resistance for the particles covered by a sur-
factant is slightly higher due to hindered charge exchange. The
Y2 parameter of CPE2, characterizing electric double-layer capac-
itance, is ≈229 for stripped TiO2 particles and 26 for as-prepared
and P25 TiO2 nanoparticles. Again, the benefits of stripped ultra-
fine TiO2 particles can be seen because the double-layer capaci-
tance is directly proportional to the interfacial area between the
electrode and the electrolyte. This is also correlated with the CV
measurements. The capacitance of the P25 sample is the same
as that of the as-prepared ultra-small TiO2 particles because the
surfactant blocks the particle’s surface, as discussed previously.
This improved capacitance was studied by CV (Figures S4
and S5, Supporting Information), with the capacitance increas-
ing from 30 F g−1 (P25); to 326 F g−1 (ultra-small TiO2 with sur-
factant); and 375 F g−1 (ultra-small TiO2 without surfactant)—
highlighting the extra available surface attained by increasing
surface area.
In order to understand and prove the phase changes detected
electrochemically, the intercalation process for TiO2 was analyzed
by using XPS in intercalated and deintercalated states (Figure 5).
To detect the presence of Ti (III) and other lower oxides of tita-
nium, pure TiO2 suspension in water was deposited onto an ITO
substrate and subjected to the same conditions as chronopoten-
tiometry with a charging current of 10 mA. A color change of
TiO2 coating to blue was observed, indicating changes in ma-
terial structure. The color starts to change back to its original
state of white, as soon as the sample is exposed back to air,
and thus the measured XPS also represents a mixture of these
blue/white states. This instability explains the low Ti3+ concen-
tration (14.8%) in Ti 2p spectra. In the intercalated stage, we can
observe the presence of lower oxidation states of titanium. The
O 1s spectra reveal the presence of three distinctive maxima at
530.7, 532.4, and 533.7 eV, which corresponds to Ti─O bonding,
─OH groups and the possible presence of adsorbed water and
organic pollutants on the sample surface respectively.
2.4. Chronopotentiometry
With a strong understanding of the electrochemical properties of
the ultra-small TiO2 electrodes, decoupled electrolysis was tested
under real conditions using chronopotentiometry. Here, a two-
electrode cell was used, where the TiO2/carbon felt electrode
(with a TiO2 loading of 7 mg cm
−2) was used as the working elec-
trode and the Pt rod was used as the auxiliary electrode. During
the chronoamperometry experiments, 10 mA of current was ap-
plied to the TiO2/carbon felt electrode until a potential of 1.5 V
was reached, with the corresponding change in potential over
time recorded, with the polarity of the applied current then re-
versed. Accordingly, OER or HER occurred on the Pt working
electrode with intercalated or deintercalated H+ ions on the aux-
iliary electrode (Figure 6a).
Figure 6b shows the first cycles of intercalation and deinterca-
lation during chronopotentiometry. When a negative potential is
measured at the TiO2/carbon felt electrode, OER occurs at the Pt
Working electrode:
2H2O→ 4H
+ + 4e− +O2 (3)
Adv. Sci. 2024, 11, 2401261 © 2024 The Authors. Advanced Science published by Wiley-VCH GmbH2401261 (5 of 10)
 21983844, 2024, 28, D
ow
nloaded from
 https://onlinelibrary.w
iley.com
/doi/10.1002/advs.202401261 by Cochrane Latvia, W
iley O
nline Library on [07/01/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on W
iley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
www.advancedsciencenews.com www.advancedscience.com
Figure 5. XPS data for TiO2 without surfactant on ITO substrates; a,c) as produced and b,d) H
+ intercalated.
And an intercalation reaction occurs at the negative potential
TiO2 auxiliary electrode.
TiO2 + xe− + xH+ → HxTiO2 + xH2O (4)
In contrast, when a positive potential is measured at the
TiO2/carbon felt electrode, a deintercalation reaction subse-
quently occurs:
HxTiO2 + xH2O→ TiO2 + xe− + xH+ (5)
with hydrogen evolution reaction on negative Pt rod:
xH+ + xe− → x∕2H2 (6)
From the shape of the chronopotentiometry curves, it can be
concluded that the TiO2 electrodes behave like supercapacitor
electrodes, as no plateau appears.[41] The time taken for intercala-
tion and deintercalation to occur, defined by the time to reach an
electrode potential of−2.05 V versusOCP and+1.5 V versusOCP
respectively, is proportional to the charge storage capacity of the
TiO2/carbon felt electrode. The differences in the intercalation
and deintercalation cycle times indicate the faradic efficiency of
the process at 93.4%.
The ultra-small TiO2-without surfactant sample demonstrates
significant storage capacity, with the first intercalation cycle tak-
ing 262 s, while the deintercalation took 260 s (Figure 6b). For the
sample TiO2 with surfactant, the times of these cycles are 47 and
43 s, respectively, and for the sample TiO2 P25, 11.2 and 10.5 s
(Figure 6b). The intercalation and deintercalation cycle time of
1000th cycles for TiO2 without surfactant lasts 328 and 269 s, re-
spectively. For TiO2 with surfactant, 47 and 44 s, while for the
TiO2 P25 sample, 15.2 and 13.3 s (Figure 6c). In situ H2 sensors
linked to a gas chromatograph were used to prove the selective
evolution of H2 gas during Step (2) (Figure S7, Supporting Infor-
mation), supporting the overall mechanism.
Over the total 1000 cycles, both the P25 and TiO2-without sur-
factant samples showed an increase in cycle duration over cycling
(Figure 6d; Figure S6, Supporting Information). A spontaneous
deintercalation process can explain this difference at longer cy-
cle times. However, the TiO2-with surfactant sample showed a
nearly 100% stability over the 1000 cycles, which is attributed to
the surfactant protecting the TiO2 particles from changing (via
aggregation or disintegration) over the reaction process.
The chronopotentiometry test was also performed at higher
currents to achieve a higher throughput of the water-splitting
reaction, approaching commercially relevant current densities
namely 50 and 100 mA. At a current of 50 mA, the TiO2-without
surfactant sample demonstrates the best performance, however,
at a current of 100 mA the charge capacity of the TiO2-with
surfactant sample is equivalent. The decrease in the capacity of
the TiO2 electrodes at higher chronopotentiometry currents is
due to the fact that a larger voltage drop occurs through the
film and less mass of the active substance participates in the
intercalation reaction. This voltage drop is higher for the TiO2-
without surfactant compared to TiO2-with surfactant, due to a
higher resistance value (Figure 4, Table 1, R1 = 2.0 vs R1 = 0.5),
thus bringing these two samples to comparable performance at
Adv. Sci. 2024, 11, 2401261 © 2024 The Authors. Advanced Science published by Wiley-VCH GmbH2401261 (6 of 10)
 21983844, 2024, 28, D
ow
nloaded from
 https://onlinelibrary.w
iley.com
/doi/10.1002/advs.202401261 by Cochrane Latvia, W
iley O
nline Library on [07/01/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on W
iley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
www.advancedsciencenews.com www.advancedscience.com
Figure 6. a) Schematic of the processes during chronopotentiometry measurements; b,c) Chronopotentiometry measurements at b) 1st cycle of all
three TiO2/carbon felt electrode samples, c) the 1000th cycle of all three samples; and d) long-term durability at different current densities (data shown
is the average of three samples for each current density).
higher current densities despite the improved surface activity of
TiO2-without surfactant. The surfactant here is thought to sta-
bilize the ultra-small TiO2 particles enabling improved rate per-
formance. The TiO2 electrodes were studied using SEM (Figure
S8, Supporting Information) before and after electrochemical cy-
cling, with no significant changes in electrode morphology being
observed.
2.5. Efficiency
To assess the efficiency of the decoupled water-splitting reaction
using TiO2 electrodes, it is necessary to quantify the amount
of hydrogen produced during step (2) of the decoupled water-
splitting reaction. To achieve this, the volumetric displacement
method, described in detail in,[15,17] was used to determine the
flow rate of the released gas. With this method, gas evolution
was determined for a sample without surfactant at a current
density of 10 mA. The time to produce 0.13 mL of H2 gas in
twelve measurements is shown in the supplementary informa-
tion Table S1 (Supporting Information). The faradaic efficiency
coefficient was also calculated for each test using the following
formula:
𝜂 = k × V × F
VM × I × t
(7)
where VM = 22.4 L mol−1 is volumetric Avogadro number, F
(C mol−1) is Faraday constant, V (L) is the displaced gas volume
per time t (s), I (A) is current and k = 2 in HER reaction.
At a current density of 10 mA, the faradaic efficiency average
of 12 measurements is 93.4%.
Knowing the faradaic efficiency, the energy efficiency was cal-
culated using the following formula:
𝜂 =
HHV ×Mmol × 𝜂F × t
F × ∫ V (t) dt
(8)
HHV is the highest heating value for hydrogen gas
(140 MJ kg−1), Mmol is hydrogen molar mass (0.001 kg mol
−1),
𝜂F is faradaic efficiency, t (s) is chronopotentiometry cycle time,
F (C mol−1) is faraday constant and integral in the denominator
is the area under the voltage curve over time. Only the deinter-
calation cycle time should be taken in nominator, as hydrogen is
produced in this cycle only. If the efficiency of the deintercalation
cycle is calculated, then only the integral of the deintercalation
cycle is taken, but if total efficiency, then both the intercalation
integral and the deintercalation.
Based on the calculations, the total energy efficiency of the
overall decoupled water-splitting reaction is 52.4%, which is
comparable to membrane electrolysis (60–80%).[45] The effi-
ciency of the deintercalation cycle reaches 167% in relation to
Adv. Sci. 2024, 11, 2401261 © 2024 The Authors. Advanced Science published by Wiley-VCH GmbH2401261 (7 of 10)
 21983844, 2024, 28, D
ow
nloaded from
 https://onlinelibrary.w
iley.com
/doi/10.1002/advs.202401261 by Cochrane Latvia, W
iley O
nline Library on [07/01/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on W
iley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
www.advancedsciencenews.com www.advancedscience.com
Table 2. Comparison of key operating parameters for current electrolyzer technologies compared to those reported in this work.[42]
Alkaline electrolyzer PEM electrolyzer Solid oxide electrolyzer This work
Operating Temperature 60–80 °C 80 °C 600–900 °C RT
Current Density Up to 400 mA cm−2 Up to 1000 mA cm−2 Up to 3200 mA cm−2 100 mA cm−2
Potential Window 1.8–2.5 V 1.7 V 1.1 V 1.5–2.0 V
Efficiency Up to 67% 80% 90% 52.4%
the consumed electricity. These calculations correspond to the
TiO2-without surfactant sample at 10 mA current (100 mA g
−1
current density) in chronopotentiometry measurements.
These efficiency values, operating temperature, voltage win-
dow and current density are competitive with current electrolyzer
systems, summarized in Table 2.[42]
As can be seen from the results, such a decoupled electroly-
sis cell is strongly asymmetric. Namely, the intercalation cycle
is extremely disadvantageous, while the deintercalation, during
whichH2 gas is produced, is advantageous. The overall efficiency
still remains at the level of efficiency of electrolyzers available in
today’s industry.[43] It is also important to compare the efficiency
of decoupled electrolysis with the conventional electrolysis effi-
ciency of the same cell. Namely, by determining the potential of
a traditional electrolysis cell at a specific current and dividing this
potential by the average potential value of the OER andHER cycle
in decoupled electrolysis mode at the same current, it is possible
to compare decoupled to traditional electrolysis.[44] The average
value of the potential in a cell with a platinum cathode and an-
ode during a period of 250 s is 2.16 V (Figure S9, Supporting
Information, shows the chronopotentiometry data). The average
potential values of HER and OER cycle are 0.94 and 1.51 V for the
P25 sample; 0.76 and 1.57 V for the sample with surfactant and
0.92 and 1.49 V for the sample without surfactant. These values
were obtained from the Figure 6b. Thus, the comparative energy
efficiency coefficients against traditional electrolysis are 88% for
the P25 sample, 90% for the sample without surfactant and 93%
for the sample with surfactant. Therefore, the energy efficiency
of decoupled electrolysis is only ≈10% behind that of traditional
electrolysis, maintaining all of the aforementioned advantages.
3. Conclusion
Herein we have demonstrated the concept of decoupled electrol-
ysis by exploiting charge storage and H+ intercalation at cheap,
abundant, pseudocapacitive TiO2 electrodes. 4.5 nm TiO2 parti-
cles deposited on carbon cloth showed an excellent pseudoca-
pacitive response of up to 375 F g−1. This high electrochemical
surface area (denoted by the high capacitance) enabled extremely
effective intercalation and deintercalation of H+ cations on the
TiO2 during decoupled water splitting. Interestingly, the bare
TiO2 electrode (without surfactant) showed impressive perfor-
mance at lower current densities (ca. 100 mA g−1), a TiO2 which
was coated with a surfactant showed equivalent performance and
greater stability at higher, commercially relevant current densi-
ties (ca. 1 A g−1). The overall efficiency of the decoupled water
splitting reaction was calculated as 52.4% approaching values for
commercial membrane electrolyzers, which operate at 80 °C in
highly alkaline conditions. Further improvements in optimizing
the efficiency of the oxygen evolution half-reaction would present
dramatic savings in energy opening up this technology for scal-
able applications.
4. Experimental Section
Synthesis of TiO2 Nanoparticles: The TiO2 nanoparticle synthesis pro-
cedure was based on a modified version of sol–gel synthesis presented by
Emmanuel Scolan and Clement Sanchez.[45,46] Titanium tetra n-butoxide
(9.05 mL, 97%, Sigma–Aldrich) was added dropwise to the mixture of
12.36 mL of n-butanol (≥99.5%, Merch, stored over CaH2) and 8.27 mL of
acetylacetone (≥99% Merck). Following this, the reaction was brought up
to reflux, and a preheated solution of 1.76 g of 4-dodecylbenzene sulfonic
acid (4-DDBSA, ≥95%, Sigma–Aldrich) in 4.865 g of deionized water was
added dropwise to the reaction mixture. After adding the 4-DBSSA, the
solution was left to reflux overnight. The formation of a clear yellow so-
lution could be observed. After cooling the mixture, a particle precipitate
was formed, which was then separated and washed with methanol (gradi-
ent grade for liquid chromatography, Supelco) using a centrifuge (2-16P,
Sigma) at 2000 × g for 1 h; after all washing cycles, particles were left to
separate at 4000 × g for 1 h and dispersed in N, N-dimethylformamide
(DMF) (99%, Merck) with fixed concentration at 100 g L−1.
For removal of surfactant, 39 mL of hexane (≥97%, Merck) was
added to 13 mL of TiO2 colloid in DMF. Thirteen milliliters of triethy-
loxonium tetrafluoroborate (EtO3BF4, ≥97%, Sigma-Aldrich) solution in
dichloromethane (DCM, for analysis, Supelco) with concentration of
20 mg mL−1 was added dropwise, followed by 20 mL of toluene (≥99.7%,
Merck). The mixture was mixed for a short period and allowed to sep-
arate. After separation, particles were washed with methanol, as stated
previously. After washing, particles were dispersed in DMF with a fixed
concentration of 100 g L−1. Electrochemical properties were evaluated for
TiO2 nanoparticles with and without surfactant.
Electrode Preparation: Electrochemical measurements were per-
formed using the ink method. TiO2 colloid was mixed with conductive
carbon black (Vulcan XC72, Nanografi) and polyvinylidene fluoride
(PVDF, average Mw 530 000, Sigma–Aldrich) in a mass ratio of 6:3:1,
respectively. Extra DMF was added to obtain suspension with the desired
viscosity. The suspension was stirred for 24 h using a magnetic stirrer and
periodical agitation using an ultrasonic bath until a homogenous mixture
was achieved. Commercially available carbon felt (AvCarb, resistance
2 Ω cm−2, fiber diameter of 20 μm) was used to produce high-capacity
electrodes. Carbon felt was washed using ultrasound for 30 min in
a mixture of sulfuric and nitric acids (volume ratio 1:2), followed by
15 min in deionized water, acetone, ethanol, and deionized water. After
all washing cycles, carbon felt samples were dried at 60 °C overnight.
Electrodes were prepared by impregnating previously washed carbon felt
with a prepared ink solution. After immersion, samples were placed on
filter paper and dried at 60 °C overnight. Samples using commercially
available TiO2 (Aeroxide P25, Thermo Fisher) were prepared similarly by
dispersing particles in DMF before adding CB and PVDF.
Sample Characterization: TiO2 nanoparticles were characterized us-
ing XRD (X’Pert, PANalytical), Raman spectrometry (inVita, Renishaw),
and transmission electron spectroscopy (TEM, Tecnai G20, FEI) imag-
ing at 200 kV. Samples in TEM were analyzed by placing them on per-
forated carbon film on a 400-mesh copper grid (S147-4, Agar Scientific).
Adv. Sci. 2024, 11, 2401261 © 2024 The Authors. Advanced Science published by Wiley-VCH GmbH2401261 (8 of 10)
 21983844, 2024, 28, D
ow
nloaded from
 https://onlinelibrary.w
iley.com
/doi/10.1002/advs.202401261 by Cochrane Latvia, W
iley O
nline Library on [07/01/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on W
iley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
www.advancedsciencenews.com www.advancedscience.com
Particle size distribution was determined by analyzing images in ImageJ
software. XPS (ESCALAB Xi+, Thermo Scientific) was used to determine
oxygen vacancies. Experimental XPS data were analyzed using Avantage 5
software, setting an advantageous carbon peak at 284.8 eV as a calibra-
tion point.[47] Electrode morphology was evaluated using scanning elec-
tron microscopy (SEM, NanoSEM 650, FEI) and energy dispersion spec-
troscopy EDS (AMETEK AplloX-SDD).
Electrochemical Measurements: Electrochemical measurements were
performed using a potentiostat (PGSTAT302N, AutoLab). The cyclic
voltammetry (CV) was performed using a three-electrode cell with a high-
capacity TiO2 electrode as a working electrode (WE), a 2 mm Pt rod
(Metrohm AG) as the reference electrode (RE), and a SCE (Hanna Instru-
ments) as the counter electrode (CE) (See Figure S1, Supporting Infor-
mation). H2SO4 (0.5 m) solution was used as an electrolyte. CVs were
performed at different scan rates from 2 to 100 mV s−1. CV curves were
analyzed using power law to consider the charge storage mechanism in
the TiO2 electrodes. Electrochemical impedance spectroscopy (EIS) was
measured in the same three-electrode setup after CV measurements at
open circuit potential (OCP). Spectra were collected with an AC signal am-
plitude of 10 mV in frequency intervals from 100 mHz to 100 kHz. The
Nyquist plot was analyzed using Nova 2 (MetrOhm) software by mod-
eling the equivalent circuit and finding the electrical parameters at the
electrode-electrolyte interface. Chronopotentiometry was measured using
a two-electrode cell with a 2 mm Pt rod as a counter electrode and refer-
ence electrode (See Figure S2, Supporting Information). TiO2/carbon felt
electrode was used as a working electrode. The electrodes were prepared
with 1 cm2 of carbon felt (specific surface area of 50 m2 g−1) loaded with
7 mg cm−2 of TiO2. For water electrolysis, the geometric area was kept
constant at 0.5 cm2 (10 x 5 mm) and the volume of the electrode was
0.15 cm3 (thickness = 3 mm). Cycling was performed using three differ-
ent current densities: 100, 500, and 1000 mA g−1 with a cut-off voltage of
1.5 and −2.05 V for 1000 cycles in ambient temperature.
Gas Chromatography–Mass Spectrometry: The identification of volatile
compounds was performed using the Gas Chromatograph–Mass Spec-
trometer (GCMS-QP2020 NX system, Shimadzu, Kyoto, Japan). The
GCMS incorporated a dielectric barrier discharge ionization detector
(BID), where samples were ionized by excited helium with an irradiation
energy of 17.7 eV. An RT-Msieve 5A column (30 m length, 0.53 mm inner
diameter, 50 μm film thickness; Restek, USA) was used for the separation
of permanent gases, with helium (grade 6.0) as the carrier gas at a lin-
ear velocity of 216.2 cm s−1. The oven was programmed to initially hold
at 35 °C for 2 min, followed by a ramp up from 35 to 100 °C at a rate of
10 °Cmin−1, with an additional holding time of 6.5 min, resulting in a total
run time of 1439 min. The injector temperature was set to 150 °C, operat-
ing in a split mode. The ion source temperature was maintained at 200 °C,
while the interface temperature was set to 190 °C. The identification of the
compounds was verified by comparing experimental retention indices to
the NIST20 spectral library.
Gas Production Rate and Efficiency Measurements: To determine the
amount of released gasses and efficiency, the cell consisting of a poly-
tetrafluoroethylene (PTFE) lid and body filled with an aqueous solution
of 0.5 m H2SO4 was used. Both electrodes and the capillary were im-
mersed in the electrolyte. The connecting tube of the medical syringe was
inserted into the cell, but its tip was above the level of the electrolyte. Dur-
ing themeasurement, an external power supply was connected to the elec-
trodes and before the potential is applied, the liquid level in the capillary
is adjusted to the lowest mark with a medical syringe. When the poten-
tial was applied, gases in the cell were being produced at 10 mA current
(100 mA g−1) and the pressure rose above the liquid level, which raised
the level in the capillary.
Supporting Information
Supporting Information is available from the Wiley Online Library or from
the author.
Acknowledgements
The authors acknowledge support from the Latvian national research
program Project No. VPP-EM-FOTONIKA-2022/1-0001 “Smart Materials,
Photonics, Technologies and Engineering Ecosystems.” P.C.S. acknowl-
edges support from RMIT University through the RMIT Vice-Chancellor’s
Research Fellowship scheme (2023).
Open access publishing facilitated by RMIT University, as part of the
Wiley - RMIT University agreement via the Council of Australian University
Librarians.
Conflict of Interest
The authors declare no conflict of interest.
Data Availability Statement
The data that support the findings of this study are available from the cor-
responding author upon reasonable request.
Keywords
electrolysis, pseudocapacitors, titanium dioxide
Received: February 2, 2024
Revised: April 30, 2024
Published online: May 14, 2024
[1] A. J. Shih, M. C. O. Monteiro, F. Dattila, D. Pavesi, M. Philips, A.
H. M. Da Silva, R. E. Vos, K. Ojha, S. Park, O. Van Der Heijden, G.
Marcandalli, A. Goyal, M. Villalba, X. Chen, G. T. K. K. Gunasooriya, I.
McCrum, R. Mom, N. López, M. T. M. W. E. Koper,Nat. Rev. Methods
Primers. 2022, 2, 84.
[2] J. Chi, H. Yu, Chin. J. Catal. 2018, 39, 390.
[3] Z. Zhu, T. Jiang, M. Ali, Y. Meng, Y. Jin, Y. Cui, W. Chen, Chem. Rev.
2022, 122, 16610.
[4] L. Fan, Z. Tu, S. H. Chan, Energy Rep. 2021, 7, 8421.
[5] J. B. Cristello, J. M. Yang, R. Hugo, Y. Lee, S. S. Park, Int. J. Hydrogen
Energy. 2023, 48, 17605.
[6] G. Zang, P. Sun, A. Elgowainy, A. Bafana, M. Wang, Environ. Sci. Tech-
nol. 2021, 55, 3888.
[7] Z. Zhu, T. Jiang, J. Sun, Z. Liu, Z. Xie, S. Liu, Y. Meng, Q. Peng, W.
Wang, K. Zhang, H. Liu, Y. Yuan, K. Li, W. Chen, JACS Au. 2023, 3,
488.
[8] Z. Zhu, Z. Liu, Y. Yin, Y. Yuan, Y. Meng, T. Jiang, Q. Peng, W. Wang, W.
Chen, Nat. Commun. 2022, 13, 2805.
[9] H. Dotan, A. Landman, S. W. Sheehan, K. D. Malviya, G. E. Shter, D.
A. Grave, Z. Arzi, N. Yehudai, M. Halabi, N. Gal, N. Hadari, C. Cohen,
A. Rothschild, G. S. Grader, Nat. Energy. 2019, 4, 786.
[10] A. Malek, X. Lu, P. R. Shearing, D. J. L. Brett, G. He, Green Energy
Environ. 2023, 8, 989.
[11] B. Rausch, M. D. Symes, L. Cronin, J. Am. Chem. Soc. 2013, 135,
13656.
[12] B. Rausch, M. D. Symes, G. Chisholm, L. Cronin, Science. 2014, 345,
1326.
[13] M. D. Symes, L. Cronin, Nature Chem. 2013, 5, 403.
[14] L. G. Bloor, R. Solarska, K. Bienkowski, P. J. Kulesza, J. Augustynski,
M. D. Symes, L. Cronin, J. Am. Chem. Soc. 2016, 138, 6707.
[15] M. Vanags, G. Kulikovskis, J. Kostjukovs, L. Jekabsons, A.
Sarakovskis, K. Smits, L. Bikse, A. Šutka, Energy Environ. Sci.
2022, 15, 2021.
Adv. Sci. 2024, 11, 2401261 © 2024 The Authors. Advanced Science published by Wiley-VCH GmbH2401261 (9 of 10)
 21983844, 2024, 28, D
ow
nloaded from
 https://onlinelibrary.w
iley.com
/doi/10.1002/advs.202401261 by Cochrane Latvia, W
iley O
nline Library on [07/01/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on W
iley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
www.advancedsciencenews.com www.advancedscience.com
[16] A. Landman, H. Dotan, G. E. Shter, M. Wullenkord, A. Houaijia, A.
Maljusch, G. S. Grader, A. Rothschild, Nature Mater. 2017, 16, 646.
[17] M. Vanags, M. Iesalnieks, L. Je¯kabsons, A. Zukuls, A. Šutka, J. Hydro-
gen Energy. 2023, 48, 20551.
[18] G. Chisholm, L. Cronin, M. D. Symes, Electrochim. Acta. 2020, 331,
135255.
[19] A. Paul, M. D. Symes, Current Opinion in Green and Sustainable Chem-
istry. 2021, 29, 100453.
[20] Communication from the Commission to the European Parliament,
the Council, the European Economic and Social Committee and the
Committee of the Regions on the 2017 List of Critical Raw Ma-
terials for the EU. https://eur-lex.europa.eu/legal-content/EN/ALL/
?uri=CELEX:52017DC0490 (accessed December 2023).
[21] Q. Peng, Z. Zhu, T. Jiang, Z. Liu, Y. Meng, S. Liu, Y. Yuan, K. Zhang, Z.
Xie, X. Zheng, J. Xu, W. Chen, ACS Appl. Mater. Interfaces. 2023, 15,
1021.
[22] J. Han, A. Hirata, J. Du, Y. Ito, T. Fujita, S. Kohara, T. Ina, M. Chen,
Nano Energy. 2018, 49, 354.
[23] J. Wang, J. Polleux, J. Lim, B. Dunn, J. Phys. Chem. C. 2007, 111, 14925.
[24] C. Chen, Y. Wen, X. Hu, X. Ji, M. Yan, L. Mai, P. Hu, B. Shan, Y. Huang,
Nat. Commun. 2015, 6, 6929.
[25] L. Mo, H. Zheng, J. Alloys Compd. 2019, 788, 1162.
[26] Y.-S. Kim, S. Kriegel, K. D. Harris, C. Costentin, B. Limoges, V. Balland,
J. Phys. Chem. C. 2017, 121, 10325.
[27] J. M. Mayer, J. Am. Chem. Soc. 2023, 145, 7050.
[28] A. Ghicov, H. Tsuchiya, R. Hahn, J. M. Macak, A. G. Muñoz, P.
Schmuki, Electrochem. Commun. 2006, 8, 528.
[29] A. Šutka, M. Iesalnieks, Titanium (iv) Oxide Based Electrod for decou-
pled electrolysis. Patent Number: LVP2024000030, 2024.
[30] R. Beranek, Adv. Phys. Chem. 2011, 2011, 786759.
[31] R. Bhola, S. M. Bhola, B. Mishra, D. L. Olson, Research Lett. Phys.
Chem. 2009, 2009, 1.
[32] S. Challagulla, K. Tarafder, R. Ganesan, S. Roy, Sci Rep. 2017, 7, 8783.
[33] H. C. Choi, Y. M. Jung, S. B. Kim, Vib. Spectrosc. 2005, 37, 33.
[34] R. Elmoubarki, F. Z. Mahjoubi, A. Elhalil, H. Tounsadi, M.
Abdennouri, M. Sadiq, S. Qourzal, A. Zouhri, N. Barka, J. Mater. Res.
Technol. 2017, 6, 271.
[35] L. Zhu, Q. Lu, L. Lv, Y. Wang, Y. Hu, Z. Deng, Z. Lou, Y. Hou, F. Teng,
RSC Adv. 2017, 7, 20084.
[36] Y. Hong, M. Yu, J. Lin, K. Cheng, W. Weng, H. Wang, Colloids Surf., B.
2014, 123, 68.
[37] R. A. Armengol, J. Lim, M. Ledendecker, K. Hengge, C. Scheu,
Nanoscale Adv. 2021, 3, 5075.
[38] W. Van Den Bergh, M. Stefik, Adv. Funct. Mater. 2022, 32, 2204126.
[39] H. Aldosari, A. Ali, S. Muqaddas, R. Shoukat, M. A. Awad, J. Mater.
Sci.: Mater. Electron. 2023, 34, 1916.
[40] S. Muthusamy, J. Charles, J. Mater. Sci.: Mater. Electron. 2021, 32,
7349.
[41] X. He, X. Zhang, J. Energy Storage. 2022, 56, 106023.
[42] C. Xiang, K. M. Papadantonakis, N. S. Lewis, Mater. Horiz. 2016, 3,
169.
[43] S. Shiva Kumar, V. Himabindu, Mater. Sci. Energy Technol. 2019, 2,
442.
[44] B.-Y. Chang, J. Electrochem. Sci. Technol. 2022, 13, 479.
[45] C. Sanchez, G. J. D. A. A. Soler-Illia, F. Ribot, D. Grosso, Comptes
Rendus. Chimie. 2003, 6, 1131.
[46] E. Scolan, C. Sanchez, Chem. Mater. 1998, 10, 3217.
[47] D. J. Morgan, C. 2021, 7, 51.
Adv. Sci. 2024, 11, 2401261 © 2024 The Authors. Advanced Science published by Wiley-VCH GmbH2401261 (10 of 10)
 21983844, 2024, 28, D
ow
nloaded from
 https://onlinelibrary.w
iley.com
/doi/10.1002/advs.202401261 by Cochrane Latvia, W
iley O
nline Library on [07/01/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on W
iley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
the Institute of Solid State Physics, University of Latvia, as the Center of Excellence has received funding from the 
European Union’s Horizon 2020 Framework Program H2020-WIDESPREAD-01-2016-2017-TeamingPhase2 under 
grant agreement No. 739508, project CAMART2.