“This document is the Accepted Manuscript version of a Published Work that appeared in final form in ACS Catalysis,
copyright © American Chemical Society after peer review and technical editing by the publisher. To access the final edited
and published work see [insert ACS Articles on Request author- directed link to Published Work, see
http://pubs.acs.org/doi/abs/10.1021/acscatal.6b01036

Neutral water splitting catalysis with a high FF triple junction polymer cell

Xavier Elias,†1 Quan Liu,†,1 Carolina Gimbert-Suriñach,†,*,2 Roc Matheu,2 Paola MantillaPerez,1 Alberto Martinez-Otero,1 Xavier Sala,4 Jordi Martorell*,1,3 and Antoni Llobet*,2,4

1

ICFO-Institut de Ciencies Fotoniques, The Barcelona Institute of Science and Technology,

08860 Castelldefels (Barcelona), Spain
2

Institute of Chemical Research of Catalonia (ICIQ), The Barcelona Institute of Science and

Technology, Avinguda Països Catalans 16, 43007 Tarragona, Spain
3

Departament de Física, Universitat Politècnica de Catalunya, Terrassa, Spain

4

Departament de Química, Universitat Autònoma de Barcelona (UAB), 08193 Cerdanyola del

Vallès, Barcelona, Spain

*

Corresponding authors e-mail: cgimbert@iciq.es, jordi.martorell@icfo.es, allobet@iciq.cat

†

These authors contributed equally to this work

The authors declare that they have no competing interests

Abstract
We report a photovoltaics-electrochemical (PV-EC) assembly based on a compact and easily
processable triple homo-junction polymer cell with high fill factor (76%), optimized conversion
efficiencies up to 8.7 % and enough potential for the energetically demanding water splitting reaction

(Voc = 2.1 V). A platinum-free cathode made of abundant materials is coupled to a ruthenium oxide on
glassy carbon anode (GC-RuO2) to perform the reaction at optimum potential (∆E = 1.70-1.78 V,
overpotential = 470-550 mV). The GC-RuO2 anode contains a single monolayer of catalyst
corresponding to a superficial concentration (Γ) of 0.15 nmol cm—2 and is highly active at pH 7. The
PV-EC cell achieves solar to hydrogen conversion efficiencies (STH) ranging from 5.6 to 6.0 %. As a
result of the solar cell’s high fill factor, the optimal photovoltaic response is found at 1.70 V, the
minimum potential at which the electrodes used perform the water splitting reaction. This allows
generating hydrogen at efficiencies that would be very similar (96%) to those obtained as if the system
were to be operating at 1.23 V, the thermodynamic potential threshold for the water splitting reaction.

Key words:
Water splitting, neutral pH, triple junction cell, OPV, GC-RuO2 anode

1-Introduction
Among all renewable energies, photovoltaics (PVs) is the most easily integrable1,2,3 in an urban
environment where most of the electricity consumption occurs and where the energy
distribution system is the largest. However, electricity production from sunlight is intermittent
and does not always match the consumption pattern. It has been proposed that an optimal
solution to perfectly combine photovoltaic integrability with a variable energy demand is to use
the excess electricity produced during peak irradiation to drive the water splitting reaction to
generate hydrogen fuel, so that the energy is stored in the form of chemical bonds. 4 Such
energy production would have a close to negligible environmental impact while its distribution
could be carried out effectively with the already existing natural gas pipelines. Unfortunately, at

2

present there are no PV hydrogen production systems that may offer integrability, low cost,
durability and a high efficiency in one.

Different PV approaches using either multi-junction or series connected cells have been
proposed to overcome the high thermodynamic potential of ∆Eo = 1.23 V of the water oxidation
reaction.5,6,7,8,9,10,11,12,13,14,15,16,17,18,19 Over-potentials associated with the reaction kinetics raise
the required voltage by several hundreds of mV to 1.5 V or higher. Early examples used
multiple junction silicon solar cells coupled to noble metal electrodes like platinum.5-7,
Although high solar to hydrogen (STH) conversion efficiencies were achieved, the prohibitive
costs of both the solar cells and the metal catalysts hampered the implementation of a large
scale hydrogen generation. Recent approaches have replaced the expensive silicon triple
junctions by metallic chalcogenide materials,10 perovskites,14-16 dye-sensitized,17,18

or

polymer18 solar cells. However, either harsh alkaline, pH=14 in the majority of them, or acidic
conditions were used to minimize the overpotentials and increase the STH efficiency of the
system. The corrosion associated to such harsh conditions prevents a practical large scale
hydrogen production using ocean or river water which is found close to a neutral pH.

In here, for an optimal water splitting reaction minimizing the overpotential limitation, we
implemented a photovoltaic-electrochemical (PV-EC) system using a one-blend polymer triple
junction cell which exhibits a high fill factor (FF) overcoming the FF of the single junction
counter parts. This implies that at the lowest measured water splitting voltage of 1.70 V, the
STH conversion efficiency under 1 sun AM1.5G illumination is as high as 96 % of the
maximum efficiency possible corresponding to an ideal operation at the thermodynamic
potential threshold. This is combined with the use of platinum free anode and cathode
electrodes immersed in a water solution buffered at pH 7. A glassy carbon rod with an
3

extremely small amount of RuO2 catalyst anchored on the graphite surface was used as anode
(GC-RuO2). In such GC-RuO2 electrode configuration a strong catalytic activity for water
oxidation is achieved at pH 7 with an amount of Ru that is at least 100 times smaller than the
Ru found in commonly used RuOx electrodes.8,18b, 20 ,21 ,22 , 23 Such a strong catalytic activity
results from an optimum electrode surface morphology that consists of a monolayer of fully
active material where all the available catalytic sites are accessible for the water oxidation
reaction. The preparation of the electrode is based on using a ruthenium molecular precursor24,25
that interconverts into the final RuO2 active species upon electrochemical treatment, details on
the synthesis and characterization of this anode can be found in Ref. [25]. We demonstrate that
a high STH conversion efficiency at pH 7 is possible when the triple junction cell and GC-RuO2
anode are combined with a cathode made of earth abundant materials deposited on a stainless
steel plate (SST-NiMoZn).11,26,27 When the described system is compared to a Pt-anode and Ptcathode electrode configuration, we obtain a 250 mV reduction in the overpotentials for
operation at pH 7. The overall PV-EC system we implement uses compact devices, earth
abundant materials or very small amounts of rare metals, operates at the pH water is found in
nature and the stability of the PV device under 1 sun illumination is ensured when a proper
encapsulation is applied.28

2-Results and Discussion
2.1-Optimal triple junction polymer solar cell
The open circuit voltage of a single junction polymer cell is not sufficient to overcome the
potential barrier for the water splitting chemical reaction. To achieve a Voc of 2.1 V we piled up
three single junction cells of the same PTB7:PC71BM blend (Fig. 1a). This blend has been
shown to exhibit one of the top single junctio29 or homo-tandem30,31 polymer cell performances.
4

By using the same blend in the three sub-cells we simplify the fabrication procedure while at
the same time we ensure a homogenous electrical performance throughout the device. This
resulted in FFs as high as 76%, significantly higher than the 71% averaged by the single
junction counterparts. Such high FFs, essential to minimize the impact of the overpotentials in
the water splitting reaction, are in correspondence to the use of very thin active blend layers
combined with a light absorption shared by the three sub-cells. In the single junction
counterpart recombination is larger given that full light absorption is carried out by just one
single junction.

Reaching the optimal light harvesting in triple junction PV architectures with high number of
layers can only be achieved by setting a target solution and implementing an inverse solving
problem approach to reach that solution.1 In particular, the cell fabricated in this work has a
total of thirteen layers with the configuration indicated in Fig. 1a. In addition, a series
connected triple junction device requires matching of the short-circuit currents of the three subcells. This current matching combined with the largest possible short circuit current is the
solution targeted by our numerical approach based on the transfer matrix32,33 to numerically
compute the electrical field distribution and light absorption within each one of the cell layers.
The optimal electrical field intensity profile and simulated EQEs for each one of the sub-cells in
the triple junction device is shown in Fig. 1b and 1c, respectively. The electric field intensity
shows maxima in the rear sub-cell for the near infrared wavelengths, in the middle sub-cell for
the wavelength range between 500 and 600 nm, and in the front sub-cell for the near UV range
(Fig. 1b). Such field distribution implies that absorption at shorter wavelengths is dominated by
the front sub-cell, in the middle range of the visible spectrum is dominated by the middle subcell, and at the near infrared by the rear sub-cell. Such perfectly balanced light absorption
distribution among the three cells is clearly seen in the simulated EQEs shown in Fig. 1c. Such
5

large differences in the EQE can be explained in part by the different thicknesses of the layers,
but also because of the interference. As can be seen in Fig. 1b constructive interference of the
incident and reflected waves tends to favor IR shifted absorption for the rear cell and UV
shifted absorption for the front cell. In one of the architectures leading to such optimal light
absorption, the PTB7:PC71BM layer thicknesses were determined to be 60 nm, 105 nm, and
110 nm for the front, middle, and rear sub-cells, respectively. These values were used to
fabricate triple-junction devices with measured conversion efficiencies up to 8.7% under AM
1.5G illumination at 100 mW/cm2 and Voc equivalent to 2.1 V. A typical current
density−voltage (J−V) curve of the triple junction cell is shown in Fig. 2, with its relevant
parameters marked in red. The potential extracted at the maximum power point of the cell was
close to 1.76 V, 530 mV higher than the thermodynamic potential for the water splitting
reaction. The high fill factor of the solar cell used (76%) ensures that the water splitting
efficiency loss due to the catalyst overpotential will be minimum as opposed to a single
junction series connection of three cells of the same polymer type that would exhibit a smaller
FF around 71% (Fig. S2), or alternative options where the Voc of the configuration would be too
large.34

2.2-Coupled GC-RuO2 anode and SST-NiMoZn cathode for water splitting catalysis
With the aim of minimizing the cost of the final PV-EC device for ultimate applications, we
used a cathode of NiMoZn26,27 on stainless steel electrode (SST-NiMoZn) made of earth
abundant materials and an anode made of RuO2 on glassy carbon electrode (GC-RuO2).25 The
selected GC-RuO2 anode is particularly interesting because it has an extremely low catalyst
loading with a superficial concentration (Γ) of 0.15 nmol cm—2 and a roughness factor (RF) of
1-2, consistent with one monolayer of active material (see supporting information). These
values are much lower than those of commonly used RuOx or IrOx based electrodes, that are
6

usually prepared using electrochemical deposition from RuCl3,20,21 dropcasting of suspensions
with post-annealing processes23 or sputtering techniques.21,23 These deposition methods give
layers of RuO2 with thickness from tens of nanometers (e.g. 70 nm in ref 21) up to a few
micrometers (e.g. 2.3 µm in ref 22). By using the molecular catalyst precursor methodology25
we have a full control of the starting monolayer of ruthenium deposited on the electrode, that
upon oxidative scans, converts into small RuO2 particles with a very high surface area available
for the water oxidation reaction. Environmental scanning electron microscopy analysis did not
show the RuO2 nanoparticles due to their small size, below the detection limit of the instrument.
However, the nature of the catalyst have been thoroughly investigated using electrochemical
and XAS techniques.25 Despite the low catalyst loading, the activity of this GC-RuO2 anode is
very high at pH 7 as demonstrated by cyclic voltammetry experiments. The catalytic wave due
to water oxidation shows a significant negative shift of the onset of the catalysis of ca. 250 mV
compared to a platinum mesh electrode (Fig. 3a).
The performance of the GC-RuO2 and SST-NiMoZn electrodes was tested independently by
means of chronopotentiometry experiments in a three electrode configuration cell at pH 7.
These experiments allowed us to simulate a real water splitting reaction that we would later do
using the triple junction organic photovoltaic cell. We fixed the current at +405 µA and –405
µA for the GC-RuO2 anode and SST-NiMoZn cathode respectively (red and pink traces, Fig.
3b). Analogous experiments were done for a platinum mesh anode and a platinum mesh
cathode for comparison (green and black traces, Fig. 3b). All electrodes are stable under
operating conditions, showing no loss of activity over the course of the experiment. It is
interesting to see that the potential required for the water reduction reaction is exactly the same
for both cathodes, the NiMoZn and Pt, accounting for less than 100 mV of overpotential, as had
been previously reported in the literature.26,27 On the other hand and consistent with the cyclic
voltammetry experiments in Fig. 3a, the performance of the GC-RuO2 is considerably superior
7

to that of platinum, which requires 250 mV more of potential than the former.
Chronopotentiometry analysis of an ITO-CoPi35,36,37,38 type electrode shows that our PV-EC
system can also operate with such cobalt based anode, but with a ∆E ca. 100 mV higher than
that of GC-RuO2 and ca. 2000 times higher catalyst loading (Γ ≈ 0.30 µmol cm-2, Fig. S1). Thus,
according to the chronopotentiometry experiments the voltage required to perform the water
splitting reaction for the GC-RuO2/SST-NiMoZn couple is ∆E = 1.78 V, which accounts for an
overpotential of 550 mV. This value is ca. 250 mV lower than that observed for the Pt/Pt or
Pt/SST-NiMoZn couples with ∆E = 2.03 (790 mV overpotential). The GC-RuO2 anode not only
minimizes the voltage at which the PV-EC device will operate, but also performs 9000
turnovers for a typical 1.5h experiment, accounting for an impressive TOF of 1.7 s-1.

2.3-Bias-free water splitting catalysis with PV-EC cell at pH = 7
Water splitting experiments were performed illuminating a triple junction solar cell (ηTJSC =
8.2-8.7%) connected to a GC-RuO2 anode and SST-NiMoZn cathode in a two electrode
configuration setup in a pH 7 phosphate buffer solution (Fig. 4a). One of the advantages of this
electrode combination is that both are highly selective, and back reactions are almost nonexistent. However, a setup configuration in a two-compartment electrochemical cell separated
by a glass frit was preferred, allowing for independent gas collection and avoiding potentially
explosive gas mixtures. The current density versus time profile of a typical reaction is given in
Fig. 4b. After an approximate 10 minutes stabilization time the current reaches a steady state at
4.5 mA/cm2, and the operating potential remains at ∆E = 1.70-1.75 V for the whole duration of
the experiment. This operating current density (JOP) in mA/cm2 can be used in

STH =

JOP · 1.23· ηF
PSolar

(1)

8

where ηF is the Faradaic efficiency and PSolar the power density, to determine the STH to be 5.6%
under 1 sun AM 1.5G illumination. As shown in Fig 5 and using the full PV-EC system,
continuous gas production was detected over a period of five hours with faradaic efficiencies
close to 100% for both gases and no apparent loss of activity.
When performing the water splitting reactions we measured working voltages in the range of
∆E = 1.70-1.80 V, which closely matched the ∆E = 1.78 V and 1.76 V calculated from the
chronopotentiometry experiments and the maximum power point of our triple junction solar cell,
respectively (Fig. 2 and Fig. 3b). The minimum potential of 1.7 V measured at some instances
with the triple junction cell relative to the chronopotentiometry measurement shown in Fig. 3b
may correspond to slight differences in the RuO2 layer coating the GC rod. The system is
sensitive to high energy light as illustrated in Fig. S3. While an initial value of STH = 6.0% is
obtained under full spectrum illumination, it decreases down to STH = 4.1% after 1.5 h. The
cell degradation induced by the UV radiation can be eliminated by placing in front of the solar
cell a 400 nm cut-off wavelength low band pass light filter. In that case we obtained stable
profiles as can be seen in Fig. 4b and Supplementary Fig. S3. These STH values are amongst
the highest reported in the literature for PV-EC hybrid devices that do not use silicon or group
III-V semiconductors as light absorber, including two recently reported perovskite-BiVO4 or
perovskite-Fe2O3 tandem assemblies (STH = 2.5 % and 2.4 % respectively),15,16 polymer solar
cell (STH = 3.1-5.4 %)18 or dye sensitized solar cell (STH = 3.1 %).17 The parameters that
allowed us to improve the latter reported benchmarking are: i) the high fill factor (76%) of the
employed triple junction cell, as opposed to the 55-60% obtained for similar configurations18b
and ii) the high performance of the GC-RuO2 electorode at pH 7, that operates at sufficiently
low overpotential. Both parameters are keys to minimize the efficiency loss during operating
conditions. Other successful PV-EC examples include series connected PV cells based on either
perovskite14 or CIGS10 semiconducting materials that provide the desired voltage and reach
9

STH higher than 10%. In these last two cases the conditions of the water splitting reaction are
either strongly alkaline (1M NaOH) for the former or highly acidic (3M H2SO4) for the latter.
In the same conditions and configuration described in Fig. 4 we performed experiments
simulating a more realistic scenario where long term stability is crucial for practical
implementation. Water splitting was performed using a larger electrochemical cell to fit higher
quantities of gases produced. After a 16 h illumination period the cells were kept in the dark for
8 h, after which time the experiment was resumed. The triple junction cell showed remarkable
stability retaining 87 % of its initial efficiency after the first experiment and 82 % after the
second one. A third cycle was performed accounting for a total illumination time of 50h and
maintaining a 79 % of the efficiency (Fig. S3 in the supporting information). The changes in the
shape of the j/E curve and its fill factor bring the optimal power point of the solar cell to lower
potentials and decrease the overall efficiency of the system; since the water splitting reaction
works at a constant potential in the range of 1.70-1.80 V, the current that passes through the
system becomes lower. However, it is remarkable to see that at the end of the 50h illumination
time all parameters still show 75% of their initial value or higher. To the best of our knowledge,
this is one of the few examples that a PV-EC device, not containing silicon or a III-V
semiconductor, shows such remarkable stability when tested for this time periods. Recent
results demonstrate that under the proper fabrication procedure, encapsulation and adequate UV
filtering solar cells based on PTB7 polymers are very stable and could improve the performance
at long time experiments. Indeed, the backbone of the polymer does not show any degradation
after more than 500 h under 1 sun continuous illumination.28

3-Conclusions

10

We have implemented a system where a 6% STH efficiency in the production of hydrogen is
achieved at neutral pH. The system relies on a simple triple junction cell with an optimal light
absorption and electrical performance. A cell FF as high as 76% allows for a water splitting
reaction at a rate (current density) which is up to 96% the rate one would achieve if water
splitting were to be carried out just above the thermodynamic potential threshold of 1.23V. In
the triple junction studied the water splitting and the maximum power point are very close
implying minimal electrical power losses. The catalysts used either for the oxidation or
reduction part of the reaction use earth abundant materials or extremely small quantities of rare
elements as Ru. To achieve a similar performance with the standard Pt electrodes one would
have to raise the pH of the water solution to 14, otherwise at pH 7 the efficiency with such
electrodes would remain close to just 3.5%. By anchoring Ru complexes on the surface of a
glassy carbon rod, we have been able to reduce by several orders of magnitude the amount of
RuO2 needed with regard to previous anode configurations using that metal oxide. In such rod
all the existing catalyst centers form a monolayer fully accessible to water molecules for
oxidation, and the catalyst is highly active at pH 7. On the other hand, the electrical power is
obtained from a PV cell based on a simple triple junction configuration where the three active
layers are based on the same PTB7 polymer blend. PTB7 has been demonstrated to be one of
the most efficient organic donor materials in photovoltaic cells. Additionally, PTB7 based cells
are optimal for the development of a PV technology with a high potential for integration.
Finally, recent work demonstrates that, contrary to results reported in prior studies, PTB7 based
cells can be made very sable when the adequate cell fabrication procedure, encapsulation, and
light filtering are applied.28
The system we report may ensure for the first time an eco-friendly large scale production of
hydrogen from water at a neutral pH. The overall system is compact and has the potential to be
fabricated at a low cost to effectively compete against other alternatives as hydrogen production
11

from methane where the cost is low but the environmental impact is large. Last but not least,
PTB7 based cells can be made semi-transparent which makes them an optimal technology for
the integration in the glass façades of city buildings where the electricity consumption is largest
and where an infrastructure for natural gas distribution, which can be effectively utilized for the
distribution of hydrogen, is already existing.

4-Experimental section
Materials. The PTB7 and PFN polymers from 1-Material, PC71BM (purity > 99%) from
American Dye Source and ZnO nanoparticles (N-10, 6083, 2.5 wt% in isopropyl alcohol) from
Nanograde were used as received. Photoactive material PTB7: PC71BM (1:1.5 wt %) were
dissolved in chlorobenzene/1, 8-diiodoctane (97:3 vol %) mixture solvent, at a total
concentration of 25 mg/mL. PFN was dissolved in methanol (1.0 mg mL-1) in the presence of
small amount of acetic acid (10 µL mL-1), which was filtered with 0.45µm PTFE filter prior to
use. ZnO solution was diluted with isopropyl alcohol (1:1, vol %) prior to use. The ZnO
formulation is very stable and the performance would not drop even when left in air for more
than one year. Ruthenium molecular precursor [Ru(bda)(NO)(N-N2)2][PF6]3 and the derivative
GC-RuO2 electrodes were prepared according to the literature, details about area and catalyst
loading can be found in the supporting information.24 The SST-NiMoZn cathodes were
prepared following a modified procedure found in the literature, details found in the supporting
information.11
Device fabrication and characterization. Triple junction devices were fabricated by spincasting ZnO solution on the pre-cleaned ITO patterned glass substrates (Lumtec, 15

/sq) and

annealing at 120 oC in glovebox for 10 min to form a 20 nm condensed electron transporting
layer. The active layer of the front sub-cell (~ 60 nm) was then spin-coated on the ZnO surface
12

and the films were dried under high vacuum (< 5 × 10−6 mbar) for one hour. Afterwards,
interconnecting layer (ICL) of MoO3 (10 nm, 0.5 Å /s)/ultrathin Ag (0.5 nm, 0.2 Å/s)/ PFN (10
nm) was deposited sequentially. Then, the active layer of the middle sub-cell (~ 105 nm) was
spin-coated on the PFN surface, dried by vacuum for one hour. The second ICL was deposited
using exactly the same procedure as the first one. The active layer of the rear sub-cell (~ 110
nm) was deposited on top of the second ICL and again left to dry in vacuum for 1 hour. Finally,
MoO3 (5 nm, 0.5 Å /s) and Ag (150 nm, 1 Å /s) electrodes were sequentially deposited through
a shadow mask by thermal evaporation (< 5 × 10− 6 mbar), which defines the device area of 9.0
mm2. The triple-junction solar cells were encapsulated with glass slides using a UV-curable
epoxy (ELC-4908, Electro-Lite Corp) in a N2 glovebox before testing in ambient air. Current–
voltage characteristics were measured under 1 sun AM 1.5G simulated sunlight (ABET Sol3A,
1000 W/m2) with a Keithley 2420. The illumination intensity of the light source (Xenon lamp,
300W, USHIO) was determined using a Hamamatsu monocrystalline silicon reference cell
calibrated by ISE Fraunhofer.

Supporting Information
Details about the equipment, materials and methods. Cathode and anode preparations.
Additional chronopotentiometry and water splitting experiments. This material is available free
of charge via the Internet at http://pubs.acs.org.

Acknowledgements: We acknowledge financial support from MINECO and the “Fondo
Europeo de Desarrollo Regional” (FEDER) through grants MAT2014-52985-R, CTQ-2013–
49075-R, SEV-2013-0319, and CTQ2014-52974-REDC. We also acknowledge financial
support from the EU COST actions CM1202 and CM1205, and the EC FP7 Program (ICT13

2011.35) under grant agreement n° NMP3-SL-2013-604506. C.G.S. is grateful to AGAUR and
GenCat for a “Beatriu de Pinós” postdoctoral grant. R.M thanks “La Caixa” Foundation for a
PhD grant and Q.L acknowledges Erasmus Mundus doctorate program Europhotonics (Grant
No. 159224-1-2009-1-FR-ERA MUNDUS-EMJD).

References
1. Betancur, R.; Romero-Gomez, P.; Martinez-Otero, A.; Elias, X.; Martorell J. Nature Photonics 2013,
7, 995-1000.
2. Zhao, Y.; Meek, G. A.; Levine, B. G.; Lunt, R. R. Adv. Opt. Mater. 2014, 2, 606-611.
3. Chen, Ch-Ch.; Dou, L.; Gao, J.; Chang, W-H.; Li, G.; Yang, Y. Energy Environ. Sci. 2013, 6, 2714.
4. Walter, M. G.; Warren, E. L.; McKone, J. R.; Boettcher, S. W.; Mi, Q.; Santori, E. A.; Lewis N. S.
Chem. Rev. 2010, 110, 6446-6473.
5. Ager, J. W.; Shaner, M. R.; Walczak, K. A.; Sharp, I. D.; Ardo, S. Energy Environ. Sci., 2015, 8,
2811-2824.
6. Khaselev, O.; Turner, J. A. Science, 1998, 280, 425-427.
7. Khaselev, O.; Bansal, A.; Turner, J. A. Int. J. Hydrogen Energy, 2001, 26, 127-132.
8. Licht, S.; Wang, B.; Mukerji, S.; Soga, T.; Umeno, M.; Tributsch H. Int. J. Hydrogen Energy 2001,
26, 653.
9. Cox, C. R.; Lee, J. Z.; Nocera, D. G.; Buonassisi, T. Proc. Natl. Acad. Sci. U.S.A. 2014, 111, 1405714061.
10. Jacobsson, T. J.; Fjällström, V.; Sahlberg, M.; Edoff, M.; Edvinsson, T. Energy Environ. Sci. 2013,
6, 3676-3683.
11. Reece, S. Y.; Hamel, J. A.; Sung, K.; Jarvi, T. D.; Esswein, A. J.; Pijpers, J. J. H.; Nocera, D. G.
Science 2011, 334, 645.
14

12. Vito, C.; Berardi, S.; Caramori, S.; Argazzi, R.; Carli, S.; Meda, L.; Tacca, A.; Bignozzi, C. Phys.
Chem. Chem. Phys 2013, 15, 13083.
13. Han, L.; Abdi, F. F.; van de Krol, R.; Liu, R.; Huang, Z.; Lewerenz, J.; Dam, B.; Zeman, M.;
Smets, A. H. M. Chem. Sus. Chem. 2014, 7, 2832-2838.
14. Luo, J.; Im, J-H.; Mayer, M. T.; Schreier, M.; Nazeeruddin, M. K.; Park, N-G.; Tilley, S. D.; Fan,
H. J.; Grätzel, M. Science 2014, 345, 1593.
15. Chen, Y-S.; Manser, J. S.; Kamat, P. V. J. Am. Chem. Soc. 2015, 137, 974-981.
16. Gurudayal, S. D.; Kumar, M. H.; Wong, L. H.; Barber, J.; Grätzel, M.; Mathews, N. Nano Lett.
2015, 15, 3833-3839.
17. Brillet, J.; Yum, J-H.; Cornuz, M.; Hisatomi, T.; Solarska, R.; Augustynsky, J.; Grätzel, M. Nature
Photonics 2012, 6, 824.
18. a) Shi, X.; Zhang, K.: Shin, K.; Ma, M.; Kwon, J.; Choi, I. T.; Kim, J. K.; Kim, H. K.; Wang, D.
H.; Park, J. H. Nano Energy, 2015, 13, 182. b) Esiner, S.; Willems, R. E. M.; Furlan, A.; Li, W.;
Wienk, M. M.; Janssen, R. A. J. Mat. Chem A 2015, 3, 23936-23945.
19. Esiner, S.; Van Eersel, H.; Wienk, M. M.; Janssen, R. A. J. Adv. Mater., 2013, 25, 2932-2936.
20. McCrory, C. C. L.; Jung, S.; Ferrer, I. M.; Chatman, S. M.; Peters, J.; Jaramillo, T. F. J. Am.
Chem. Soc. 2015, 137, 4347.
21. Tsuji, E.; Imanishi, A.; Fukui, K.-I.; Nakato, Y. Electrochimica Acta. 2011, 56, 2009-2016.
22. Licht, S.; Wang, B.; Mukerji, S. J. Phys. Chem. B 2000, 104, 8920-8924.
23. Spurgeon, J.; Velazquez, J. M.; McDowell, M. T. Phys. Chem. Chem. Phys. 2014, 16, 3623-3631.
24. Duan, L.; Bozoglian, F.; Mandal, S.; Stewart, B.; Privalov, T.; Llobet, A.; Sun, L. Nature Chem.
2012, 4, 418-423.
25. Matheu, R.; Francàs, L.; Chernev, P.; Ertem, M. Z.; Batista, V.; Haumann, M.; Sala, X.; Llobet, A.
ACS Catalysis. 2015, 5, 3422.
15

26. Stachurski, J. Z. O.; Williamsville, D. P.; Ripa, J. A.; Pokrzyk, G. F. U. S. Patent 1982, 4,354,915.
27. Nocera, D. G. Acc. Chem. Res. 2012, 45, 767-776.
28. Liu, Q.; Mantilla-Perez, P.; Montes Bajo, M.; Romero-Gomez, P.: Martorell, J. to be published.
29. Li, G.; Zhu, R.; Yang, Y. Nature Photon. 2012, 6, 153-161.
30. Bahro, D.; Koppitz, M.; Mertens, A.; Glaser, K.; Mescher, J.; Colsmann, A. Adv. Energy Mater.
2015 DOI: 10.1002/aenm.201501019.
31. Martínez-Otero, A.; Liu, Q.; Mantilla-Perez, P.; Montes Bajo, M.; Martorell, J. J. Mater. Chem. A
2015, 3, 10681-10686.
32 . Mantilla-Perez, P.; Martínez-Otero, A.; Romero-Gomez, P.; Martorell, J. ACS Appl. Mater.
Interfaces 2015, 7, 18435-18400.
33. Martínez-Otero, A.; Elias, X.; Betancur, R.; Martorell, J. Adv. Optical Mater. 2013, 1, 37-42.
34. Urbain, F.; Smirnov, V.; Becker, J-P.; Lambertz, A.; Rau, U.; Finger, F. Sol. Energ. Mater. Sol.
Cells 2016, 145, 142–147.
35. Kanan, M. W.; Nocera, D. G. Science 2008, 321, 1072-1075.
36. Lutterman, D. A.; Surendranath, Y.; Nocera, D. G. J. Am. Chem. Soc. 2009, 131, 3838-3839.
37. Esswein, A. J.; Surendranath, Y.; Reece, S. Y.; Nocera, D. G. Energy Environ. Sci. 2011, 4, 499504.
38. Zhong, D. K.; Choi, S.; Gamelin, D. R. J. Am. Chem. Soc. 2011, 133, 18370-18377.

16

Figures

Figure 1 Optically optimal triple junction cell a) Schematic representation of the triple junction
solar cells and molecular structure of its organic polymers. PTB7, Poly({4,8-bis[(2ethylhexyl)oxy]benzo[1,2-b:4,5-b′]dithiophene-2,6-diyl}{3-fluoro-2-[(2ethylhexyl)carbonyl]thieno[3,4-b]thiophenediyl}. PC71BM, [6, 6]-phenyl-C71 butyric acid
methyl ester. PFN, poly [(9,9-bis(3'-(N,N-dimethylamino)propyl)-2,7-fluorene)-alt-2,7-(9,9–
17

dioctylfluorene. b) Calculated electrical field intensity normalized with respect to the incident
field intensity as a function of wavelength for an optimized triple junction organic solar cell
architecture. c) Calculated EQE for each one of the three sub-cells in an optimized architecture.

Figure 2. j/E curve of triple junction cell. VOC = 2.13 V, JSC = 5.4 mA/cm2, FF = 76% and ηmax
= 8.7% where the water splitting thermodynamic potential (orange) and water splitting
operating potential range (green) are indicated. VOC: open circuit potential, JSC: short circuit
current density, FF: fill factor and ηmax = cell efficiency at maximum power point.

18

Figure 3. a) Cyclic voltammetry experiments of a GC-RuO2 (red) and platinum mesh (green)
working electrodes. b) Chronopotentiometry experiments at a fixed current of +405 µA for GCRuO2 (red) and Pt mesh (green) or –405 µA for NiMoZn (pink) and Pt mesh (black). All
experiments in a) and b) were done in a two compartment cell, in degased 0.1 M phosphate
buffer, at pH 7.0, using a platinum mesh as counter electrode and Ag/AgCl as reference
electrode. The potentials are given against the NHE reference electrode using the conversion
ENHE = EAg/AgCl + 0.2. The thermodynamic potentials for the water oxidation (E0ox, pH 7 = 0.817
V) and water reduction (E0red, pH 7 = –0.413 V) reactions at pH 7 are indicated in grey dotted
lines.
19

Figure 4. Water splitting EC-PV system a) Schematic representation of the setup for the
water splitting experiments using a triple junction organic solar cell, GC-RuO2 anode
and SST-NiMoZn cathode. b) Current vs time profile of a water splitting experiment
using a triple junction solar cell (ηTJSC = 8.5 %), GC-RuO2 anode and SST-NiMoZn
cathode in a two electrode configuration and a two compartment cell containing 0.1 M
phosphate buffer, pH = 7.0, under AM 1.5 G illumination with a GG400 filter. The
operating potential of the reaction was measured manually and is also indicated. After
1.5 h, the amount of evolved gas accounted for TONRu = 9 000 and TOF = 1.7 s-1, where
TONRu = mols O2/mols of Ru and TOF = mols O2/(mols of Ru×second).

20

Figure 5. Gas evolution of a water splitting experiment using a triple junction solar cell
(ηTJSC = 8.2%), GC-RuO2 anode and SST-NiMoZn cathode in a two electrode
configuration and a two-compartment cell, containing 0.1M phosphate buffer, pH = 7,
under 1 sun irradiation (λ > 400 nm). The dotted lines represent the 100% faradaic
efficiency based on the charge passed during electrolysis while the solid lines represent
the gas measured with Clark electrode sensors.

21

Table of Contents and Abstract Graphic

ToC Text:
Efficient visible light induced hydrogen production from water at neutral pH, using a simple triple
junction cell coupled to a RuO2 anode and a NiMoZn cathode, reaching 6% STH.

22