Nickel-Based Dye-Sensitized Photocathode: Towards Proton Reduction Using a Molecular Nickel Catalyst and an Organic Dye

Article abstract of DOI:10.1002/cctc.201600025

To construct an efficient dye-sensitized photo-electrochemical tandem cell for hydrogen production, it is crucial to understand the working principles of both the photoanode and the photocathode. Herein, the anchoring of a proton-reduction catalyst and an

Full text of DOI:10.1002/cctc.201600025

DOI: 10.1002/cctc.201600025
Full Papers
Nickel-Based Dye-Sensitized Photocathode: Towards
Proton Reduction Using a Molecular Nickel Catalyst and an
Organic Dye
Bart van den Bosch,[a] Jeroen A. Rombouts,[b] Romano V. A. Orru,[b] Joost N. H. Reek,*[a] and
Remko J. Detz*[a]
To construct an efficient dye-sensitized photo-electrochemical
tandem cell for hydrogen production, it is crucial to under-
stand the working principles of both the photoanode and the
photocathode. Herein, the anchoring of a proton-reduction
catalyst and an organic dye molecule on metal oxides is stud-
ied for the preparation of a photocathode. On TiO2, the Ni cat-
alyst behaves as a good electrocatalyst (À250 mAcmÀ2) in
acidic water (pH 2). The Ni catalyst and the organic dye were
co-immobilized on NiO to form a solely Ni-based photoca-
thode. The electron-transfer steps were investigated by using
various techniques (IR, UV/Vis, and fluorescence spectroscopy,
and (photo)electrochemistry). Despite the observed successful
single-electron-transfer steps between all of the components,
photocatalysis did not yield any hydrogen gas. Possible bottle-
necks that prevent photocatalytic proton reduction are poor
electron transfer because of aggregation, charge recombina-
tion from the catalyst to the NiO, or instability of the catalyst
after the first reduction.
Introduction
The storage of solar energy as fuel by the light-driven splitting
of water into oxygen and hydrogen is one of the most promis-
ing strategies to fulfill the future need for energy in a sustaina-
ble way. Photo-electrochemical cells (PECs) are attractive in this
respect because the evolution of oxygen and hydrogen occurs
at the physically separated anode and cathode, which simpli-
fies the isolation of the gasses. Although several PECs have
been reported that split water using inorganic light-harvesting
materials (e.g., a-Si, BiVO4, GaInP2), the tuning of the bandgaps
and band edges of these materials remains difficult and gener-
ally the stability of devices in water is poor.[1]
that performs the chemical transformation. Initially, in analogy
to dye-sensitized solar cell research, DS-PEC research focused
mainly on the development of a photoactive anode for water
oxidation using Pt as the catalytic cathode for proton reduc-
tion.[2] Although light-driven proton reduction on a solid sup-
port has been reported,[3] only a few examples elaborate on
the preparation of a photoactive cathode for proton reduction
using NiO as the semiconductor material.[4]
Remarkably, no examples exist of the use of one of the fast-
est molecular proton-reduction catalysts (PRC), a Ni-DuBois-
type PRC,[5] on a NiO electrode. In 2012, Li et al. described for
the first time visible-light-driven hydrogen generation using an
organic dye-modified NiO cathode in combination with a mo-
lecular Cobaloxime PRC.[6] In this system, the photocurrent de-
creased rapidly upon irradiation, which is assigned to either
the decomposition of the catalyst or the leaching of the cata-
lyst into solution. To improve catalyst binding, Ji et al. reported
a photocathode in which a Ru-dye molecule is anchored on
NiO and to the Cobaloxime catalyst through a supramolecular
interaction.[7] With this system, the photocurrent was stable for
several hours. Sun et al. combined these photocathodes for
the first time with a photoanode that formed the first DS-PEC
that could split water in a two-electrode setup.[8]
Dye-sensitized photo-electrochemical cells (DS-PECs) use
molecular dye molecules as light-harvesting compounds. The
absorbance, redox properties, and stability of these dye mole-
cules can be optimized by synthetic modification, which is
a major advantage for the optimization of DS-PECs. In DS-
PECs, typically, a wide-band-gap semiconductor material is
coated with dye molecules that are also coupled to a catalyst
[a] Dr. B. van den Bosch, Prof. Dr. J. N. H. Reek, Dr. R. J. Detz
Van’t Hoff Institute for Molecular Sciences
University of Amsterdam
Science Park 904
1098 XH Amsterdam (The Netherlands)
E-mail: j.n.h.reek@uva.nl
Herein, we report the construction of a NiO photocathode
modified with an organic dye molecule and a Ni-DuBois-type
PRC to form a new Ni-based photoelectrode.
r.j.detz@uva.nl
[b] J. A. Rombouts, Prof. Dr. R. V. A. Orru
Division of Organic Chemistry
Vrije Universiteit Amsterdam
De Boelelaan 1083a
1081 HV Amsterdam (The Netherlands)
Supporting Information and the ORCID identification number(s) for the
author(s) of this article can be found under http://dx.doi.org/10.1002/
cctc.201600025.
ChemCatChem 2016, 8, 1392 – 1398
1392
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
Catalyst in solution
Results and Discussion
First, Ni complex 1 (Ni-1) was synthesized and studied by elec-
trochemistry in solution. The cyclic voltammogram of Ni-1 in
CH3CN/MeOH (4:1) displays one irreversible reduction at
À0.9 V (vs. ferrocene/ferrocenium; Fc/Fc+), which is ascribed to
the NiII/NiI redox couple. This is in contrast to most reported
bisdiphosphine nickel PRCs, which have two consecutive rever-
sible reduction waves that correspond to the NiII/NiI and NiI/Ni0
redox couples.[16]
Electrode design
As demonstrated previously for p-type dye-sensitized solar
cells (DSSC),[9] NiO is also a promising wide-bandgap semicon-
ductor material for photocathode preparation for DS-PECs. The
valence band (VB) potential energy of around 0.68 V vs. the
normal hydrogen electrode (NHE; at pH 2)[10] determines the
maximum reduction potential of the excited state of the dye
to allow the quenching of the excited dye molecule (Figure 1).
The irreversibility of the reduction and the absence of
a second reduction might be caused by the poor solubility of
the one-electron reduced species. A similar observation was re-
ported for a bisdiphosphine nickel complex functionalized
with carboxylic acids.[17] The addition of protonated DMF
(DMFH·OTf) renders a catalytic reduction wave around À0.8 V
(vs. Fc/Fc+) in the cyclic voltammogram (Figure 2), which is as-
cribed to catalytic proton reduction. This potential is expected
to be sufficiently low to be driven by potentials at the level of
the CB of TiO2. In the following section the immobilization of
Ni-1 on a TiO2 electrode is described.
Figure 1. Schematic representation of the design parameters for the con-
struction of a NiO-photocathode modified with an organic dye molecule
and a Ni-DuBois-type PRC.
Figure 2. Cyclic voltammograms of a solution of Ni-1 (black line) and a solu-
tion of Ni-1 in the presence of increasing amounts of DMFH·OTf (2, 3, 5, 7, 8,
and 10 equiv., gray lines). Conditions: 0.1m Bu4NPF6 in CH3CN/CH3OH (2:1),
mercury droplet working electrode, Ag wire counter electrode, scanning
speed 0.1 VsÀ1, referenced to Fc/Fc+.
In addition, the reduced dye molecule should be able to
reduce the PRC. For this purpose, a naphthalene diimide (NDI)
scaffold was chosen as the core of our organic dye molecule.
These NDIs have a typical HOMO level of 0.88 V vs. NHE.[11] This
should make the quenching of the excited state of the dye by
NiO possible. The reduction potential of the dye (À1.16 V vs.
NHE) is sufficient to drive a Ni-DuBois-type PRC, which is re-
ported to be active at a low overpotential in water.[3a,12]
Catalyst on TiO2
Ni-1 is immobilized on fluorine-doped tin oxide (FTO) coated
glass slides that bear a mesoporous layer (10 mm) of TiO2 nano-
particles (20 nm). This allowed us to study this immobilized
catalyst in the electrocatalytic reduction of protons in water.[18]
After dip-coating the TiO2-electrode in a solution of Ni-
1 (0.15 mm in CH3CN/MeOH, 2:1) for 4 h, the now yellow-col-
ored electrode was, after thorough washing, analyzed by using
attenuated total reflectance (ATR) FTIR spectroscopy (see Sup-
porting Information). Peaks that correspond to both Ni-1 and
TiO2 appear in the ATR-FTIR spectrum of the Ni complex
coated on the TiO2 electrode (Ni-1@TiO2) electrode. Further-
more, inductively coupled plasma optical emission spectrosco-
py (ICP-OES) of Ni (after the removal of the catalyst from the
A co-immobilization strategy is sought to simplify the syn-
thetic scheme as the dye and catalyst can be prepared sepa-
rately. The same approach was used to prepare the NiO-based
photocathodes reported previously[7,8] and, moreover, to make
photoanodes for water oxidation.[13] These studies already
demonstrated that the choice of the anchoring group[14] and
the distance[15] between the anchored components is very im-
portant for the electrode performance. To achieve a strong
binding of the molecules to the NiO surface, hydroxamic acids
were chosen as anchoring groups, which have been reported
to bind very strongly to metal oxide surfaces.[14d]
ChemCatChem 2016, 8, 1392 – 1398
www.chemcatchem.org
1393
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
surface in nitric acid) revealed that the catalyst loading was
76 nmolcmÀ2.
The electrocatalytic activity was studied by submerging the
Ni-1@TiO2 electrode in an aqueous phosphate buffer (purged
with N2, 0.1m, pH 2.1) in a three-electrode setup using a Pt coil
as the counter electrode and a Ag/AgCl reference electrode. At
a bias potential of À0.65 V, bubble formation was observed,
and analysis of the headspace by GC proved the formation of
hydrogen. After equilibration, a stable current density of
À250 mAcmÀ2 was observed (Figure S4), which after half an
hour decreased slowly to zero (after 1 h, 54 mLH cmÀ2 was
2
formed with a turnover number (TON) of 32). ATR-FTIR spec-
troscopy after electrocatalysis revealed that the catalyst was no
longer bound to the electrode surface. The Faradaic efficiency
was calculated to be 62% (after 1 h). This was lower than that
described by others for a similar Ni catalyst in water at pH 4.5
(85%),[3a] which can be ascribed partly to the charging of the
TiO2.
Figure 3. ATR-FTIR spectra of Ni-1@NiO+Dye-2, Dye-2@NiO+Ni-1, Ni-1@NiO,
and Dye-2@NiO. The order indicates the steps of immobilization, for exam-
ple, first immobilization of complex, followed by immobilization of the dye
is Ni-1@NiO+Dye-2.
The stability was improved by exchanging the phosphate
buffer for a trifluoroacetate (TFA) buffer (0.1m pH 2.1), al-
though at the expense of a lower hydrogen production (after
1 h, 15 mL H2 cmÀ2 was formed, TON=9) and current density
(À60 mAcmÀ2 after 1 h; Figure S5). Importantly, 50% of the ini-
tial activity was still observed after an impressive 3 h of electro-
catalysis, although with a low Faraday efficiency (46%). These
results were sufficiently convincing to proceed to the next
step, which is the immobilization of Ni-1 on NiO together with
the organic dye molecule to construct the NiO photocathode.
solution that contained Ni-1. Slight differences in catalyst load-
ing were measured: 4.3 nmolcmÀ2 for Dye-2@NiO+Ni-1 and
5.2 nmolcmÀ2 for Ni-1@NiO+Dye-2 (ICP analysis of phospho-
rus, see also Supporting Information). The ATR-FTIR spectrum
of the Ni-1@NiO+Dye-2 electrode (the order indicates the im-
mobilization steps, for example, first immobilization of the
complex followed by immobilization of the dye gives: Ni-
1@NiO+Dye-2) shows the indicative n(C=O) signals of the hy-
droxamic acids. The peaks around n˜ =1603 cmÀ1 (indicated
*
with ) correspond to the C=O bond of the hydroxamate of
the catalyst. The peaks at n˜ =1580 cmÀ1 (indicated with N)
correspond to the C=O bond of the hydroxamate of the pho-
tosensitizer. This shows clearly that various ratios of dye and
catalyst can be obtained by changing the loading conditions.
The functionalized electrodes were further characterized by
using UV/Vis spectroscopy with integrated-sphere diffuse re-
flectance technology (Figure 4). The absorption of the immobi-
lized Dye-2 was observed clearly, which indicates successful
immobilization. The relative intensity of the two major absorp-
tions was changed upon immobilization. In CH2Cl2, the absorp-
tion at l=599 nm was more intense than that at l=552 nm.
After binding on the NiO surface, the absorption at l=650 nm
became more intense than that at l=615 nm. This change in
the shape of the absorption spectrum is indicative of aggrega-
Catalyst and dye on NiO
The dye molecule 2 (Dye-2) was synthesized according to
a slight modification of a literature procedure.[11,19] Next, Dye-2
was anchored onto FTO-coated glass slides that bear a meso-
porous layer of NiO nanoparticles by dipping the electrode
into a solution of Dye-2 in CH2Cl2 (0.6 mm). After washing, the
obtained dark purple electrodes were analyzed by ATR-FTIR
spectroscopy, which revealed several peaks that correspond to
the dye molecule next to the peaks of NiO. The reduction po-
tential of the dye on NiO was À1.16 V vs. NHE if measured in
CH2Cl2 (Figure S8), no data could be obtained in acidic water
because of a high background current. The immobilization of
Ni-1 on NiO was performed in a similar fashion as that of TiO2
and gave also similar results as illustrated by the ATR-FTIR
spectra (Figure 3 and Supporting Information). The co-immobi-
lization of Ni-1 with Dye-2 was achieved by submerging the
Ni-1@NiO electrode in a solution that contained Dye-2. Alter-
natively, the Dye-2@NiO electrode can be submerged in the
Figure 4. Reflectance UV/Vis spectra of NiO, Dye-2@NiO, and Dye-
2@NiO+Ni-1 and UV/Vis spectrum of Dye-2 in CH2Cl2.
ChemCatChem 2016, 8, 1392 – 1398
www.chemcatchem.org
1394
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
tion (or excimer formation), which is reported for various pery-
lene diimide derivatives in solution.[20]
we describe the electrochemical experiments that were con-
ducted to see if the energy levels of the immobilized compo-
nents are still properly aligned for electron transfer to take
place.
Furthermore, after the immobilization of Dye-2, the absorp-
tion was broadened dramatically compared to that of the dye
in solution, and a redshift of all the maxima in the absorption
spectra was observed. These two phenomena have been re-
ported previously for the immobilization of other dyes on NiO
and are ascribed to the coupling of the PS to NiO.[21] The
broadening of the absorptions is more profound with the Dye-
2@NiO electrodes than for the Dye-2@NiO+Ni-1 electrode. In
the latter, the co-immobilization of Ni-1 probably separates the
dye molecules, which thereby reduces aggregation. This sug-
gests that the redshift and broadening of the absorptions and
the change in the band shape is caused mainly by the aggre-
gation of Dye-2 and to a lesser extent by the coupling of the
dye with NiO.
Photo-electrochemistry
Linear sweep voltammetry measurements were performed
with the NiO and the Dye-2@NiO electrodes as the working
electrodes. We scanned the NiO electrodes and the Dye-2@NiO
electrodes oxidatively in acetate buffer pH 4.7, while shielding
them from light, and practically no current was observed be-
tween À0.2 and 0.15 V. Upon shining light on the NiO elec-
trode no increase in current was observed, whereas shining
light on the Dye-2@NiO electrode resulted in a photocurrent
of À40 mAcmÀ2 (Figure S11). This current arises most probably
from the reduction of oxygen by the reduced dye molecule as
almost no photocurrent was observed under oxygen-free con-
ditions (Figure S12). Importantly, this indicates that the VB of
NiO is properly aligned with the reduction potential of the ex-
cited dye and that photoinduced hole injection into the NiO
and hole transport through the nickel oxide occurs upon the
excitation of Dye-2. The gradual increase in current at 0.15 V is
ascribed to oxidation of the valence band of NiO. Furthermore,
we used amperometry at a bias voltage of À0.1 V vs. Ag/AgCl
in phosphate buffer pH 2, to measure that the Dye-2@NiO
electrode gave a photocurrent of À12 mAcmÀ2 during two
consecutive light on/off cycles (Figure 6) without any decrease
Additionally, the emission of the immobilized Dye-2 after ex-
citation at l=565 nm is redshifted and strongly broadened
compared to that of Dye-2 in CH2Cl2, which indicates aggrega-
tion (Figure 5). To our surprise, the emission of Dye-2@NiO+Ni-
Figure 5. Fluorescence spectra (solid-state reflectance) of NiO, Dye-2@NiO,
Dye-2@NiO+Ni-1, and (in solution) Dye-2 in CH2Cl2.
1 is much higher than that of Dye-2@NiO.[22] An explanation
for the increase in emission for the Dye-2@NiO+Ni-1 electrode
is that aggregation of the dye in the Dye-2@NiO electrode has
a detrimental effect on the fluorescence. Unfortunately, be-
cause the quantum yield of the immobilized dye could not be
determined, the percentage of dye that emits fluorescence is
unknown, and the expected fluorescence quenching by hole
injection into the NiO or electron transfer to the catalyst
cannot be quantified in this way. Moreover, the dye is not
quenched fully by hole injection as emission was still ob-
served.
Figure 6. Amperometry of the Dye-2@NiO electrode in air (solid line) and
flushed with N2 (dotted line). Conditions: 0.1m phosphate buffer pH 2, Pt
counter electrode, leak-free Ag/AgCl reference electrode, AM1.5G solar simu-
lation, 100 mWcmÀ2 with 408 nm longpass filter, bias potential of À0.1 V vs.
Ag/AgCl.
in photocurrent, which indicates that the system is stable
under the applied conditions. Similar observations were made
if we used sodium acetate buffer at pH 4.7 (Figure S13).
Based on the results of ATR-FTIR, fluorescence, and reflec-
tance UV/Vis spectroscopy, we conclude that both the dye and
catalyst were immobilized successfully and co-immobilized on
the NiO surface. Both the absorption and emission spectra of
Dye-2 changed upon immobilization. This means that the ex-
cited state energy of the immobilized dye also deviates from
the excited state energy in solution. In the following section,
To see if photoinduced electron transfer can also take place
from Dye-2 to the Ni-1 PRC, photocurrent measurements
under anaerobic conditions were performed with Ni-1 dissolved
in a mixture of methanol/acetonitrile (1:1). In the first attempts,
the photocurrent was not stable and decreased rapidly (a few
ChemCatChem 2016, 8, 1392 – 1398
www.chemcatchem.org
1395
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
minutes) because of the replacement of the dye molecules by
the catalyst, which was observed by ATR-FTIR spectroscopy.
The experiment was repeated in acetonitrile using Ni-3,[16]
which lacks the hydroxamic acid binding moieties. Now
a stable photocurrent was observed for several minutes during
amperometry. The current increased if we increased the con-
centration of Ni-3, which demonstrates that Ni-3 accepts elec-
trons from the reduced dye and acts as a redox mediator
(Figure 7). A similar electron-transfer chain from NiO to the cat-
alyst was also observed by Odobel et al.[4b]
or by the instability of the catalyst after the first reduction to
prevent the final step to achieve proton reduction.
Currently, our research focuses on how to achieve light-
driven proton reduction on surfaces by slowing down recombi-
nation processes and improving catalyst activity and stability.
To avoid aggregation issues, strategies are conducted to
couple the catalyst to the dye molecules and prepare dyad
structures, which can be immobilized in the assembled form.
Conclusions
Proton-reduction catalyst Ni-1 could be attached to TiO2 and
was found to be a good electrocatalyst in aqueous solution
(pH 2.1) to give a stable current of À250 mAcmÀ2 (overpoten-
tial of ꢀ300 mV). The catalyst could also be immobilized to-
gether with Dye-2 on NiO electrodes to construct a Ni-based
photocathode. It was shown that electron transfer from the
NiO to the dye molecule and from the dye to the catalyst can
occur upon light irradiation. Unfortunately, the electrodes are
not photocatalytically active under irradiation in various acidic
conditions upon application of a bias voltage of À0.1 V vs. Ag/
AgCl. Possible bottlenecks that prevent photocatalytic proton
reduction are poor electron transfer because of the aggrega-
tion of the molecules on the surface, charge recombination
from the catalyst to the NiO, or the instability of Ni-1 after the
first reduction.
Figure 7. Amperometry of the Dye-2@NiO working electrode of various con-
centrations of Ni-3 in CH3CN purged initially with N2. Conditions: 500 W
XeHg lamp with 470 nm longpass filter, Pt coil as counter electrode, bias po-
tential of À0.1 V vs. Ag/AgCl.
Experimental Section
Synthesis of Ni-1
Photocatalysis
Synthesis of 4-amino-N-hydroxybenzamide: NaOH (3.2 g,
80 mmol) and NH2OH·HCl (2.78 g, 40 mmol) were dissolved in
water (50 mL). Methyl 4-aminobenzoate (2.00 g, 13 mmol) dis-
solved in CH3OH (30 mL) was added. This mixture was stirred at RT
for 2 days, after which it was neutralized to pH 7 with 6m HCl. The
resulting solution was extracted with ethyl acetate (3”40 mL). The
combined organic layers were dried over MgSO4 and concentrated
till some white precipitate was formed. This solution was placed at
48C overnight. The resulting white precipitate was collected by fil-
As we knew that all of the single-electron-transfer steps can
occur, the photocatalytic properties of the photocathodes
were evaluated in a standard three-electrode setup, with Ag/
AgCl as the reference electrode and a Pt coil as the counter
electrode. The irradiation of both the Dye-2@NiO+Ni-1 and
the Ni-1@NiO+Dye-2 electrodes with light (l>470 nm) in
phosphate buffer (pH 2.1) with a bias potential of À0.1 V did
not lead to any photocurrent, and headspace analysis by GC
did not reveal the formation of hydrogen gas. Unfortunately,
also in experiments in which we used a sodium acetate buffer
(pH 4.7) or 10 mm DMFH+ in acetonitrile no observable photo-
current or hydrogen formation was observed.
1
tration and dried in vacuo. Yield: 54%. H NMR ([D6]DMSO, 298 K):
d=10.75 (br s, 1H), 8.67 (br s, 1H), 7.47 (d, J=8.6 Hz, 2H), 6.53 (d,
J=8.6 Hz), 5.62 ppm (br s).
Synthesis of the ligand: Phenylphosphine (145 mL, 1.3 mmol) and
formaldehyde (purged with N2, 35% in water, 235 mL, 2.9 mmol)
were heated to reflux in ethanol (5 mL; purged with N2) for 18 h,
after which all volatiles were removed in vacuo. The resulting
white oil was dissolved in ethanol (5 mL; purged with N2) and 4-
amino-N-hydroxybenzamide (200 mg, 1.3 mmol) dissolved in water
(7 mL; purged with N2) with heating was added to the solution.
The mixture was heated to reflux for 18 h. Another portion of 4-
amino-N-hydroxybenzamide (100 mg, 0.66 mmol) dissolved in
water with heating (3 mL; purged with N2) was added to the solu-
tion and the mixture was heated to reflux for another hour to give
full conversion (followed by 31P NMR spectroscopy). The resulting
white precipitate was collected by filtration over a glass filter,
washed with water, and the white solid dried in vacuo. Yield: 54%.
1H NMR ([D6]DMSO, 258C) peaks of the major (83%) compound
(see also Supporting Information) d=10.81 (s, 2H), 8.74 (s, 2H),
7.72 (m, 4H), 7.58 (d, J=8.9 Hz, 4H), 7.54 (m, 4H), 7.44 (m, 2H),
The co-immobilization strategy allows the aggregation of
the dye (or catalyst) molecules with each other, as was indicat-
ed by UV/Vis and fluorescence spectroscopy on the electrodes,
instead of the formation of a homogeneous mixture of the
molecules on the surface. This aggregation can prevent effi-
cient electron transfer between reduced dye molecules and
catalysts. Before the reduction of the catalyst takes place, a re-
combination event may occur because of electron injection
from the reduced dye back into the NiO, which can explain
the lack of photocatalytic activity.[23] Another possibility is that
the second reduction of the co-immobilized Ni-1 does not
happen, which thus prevents photocatalysis in these cathodes.
This can be explained by slow electron delivery to lead to fast
recombination from the reduced catalyst to the oxidized NiO[3c]
ChemCatChem 2016, 8, 1392 – 1398
www.chemcatchem.org
1396
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
6.71 (d, J=8.9 Hz, 2H), 4.64 (m, 4H), 4.20 ppm (m, 4H); 31P NMR
([D6]DMSO, 258C): d=À49.5 ppm (s); IR: n˜ =1600 (s), 1481 (s),
1454, 1434, 1383, 1333, 1307, 1250, 1211 (s), 1147 (s), 1025, 976,
896, 815, 751 (s), 692 (s), 602, 518, 469, 416 cmÀ1; HRMS (ESI): m/z:
calcd for [C30H30N4O4P2Na]+: 595.1640; found: 595.1603.
extracted with 3”200 mL CH2Cl2. The combined organic layers
were dried over MgSO4 and concentrated in vacuo, after which
column chromatography (alumina; eluent system 1:1 pentane/
CHCl3) was performed to yield the product as an orange-yellow
solid. Yield: 8%. NMR spectroscopic data of the compound is iden-
tical to literature values.[25] M.p.: 233–2358C (decomp); TLC
Synthesis of Ni-1: The ligand (44 mg, 0.076 mmol) and
Ni(BF4)·6CH3CN (18 mg, 0.038 mmol) were heated in CH3CN (6 mL)
at 508C. After 1 h, CH3OH (1.5 mL) was added, and the red/white
suspension was heated to reflux for 1.5 h. At this point, a solution
of HBF4·ether (5.5 mL in 1 mL CH3CN) was added, and the mixture
was stirred for 1 h upon which it turned bright red and the white
precipitate had almost disappeared. After filtration and evapora-
tion of the solvent a red solid was obtained, which was dried in
1
(CH2Cl2): Rf =0.7, yellow spot; H NMR (500 MHz, CDCl3, 258C): d=
3
4.19 (d, 4H, J=7.5 Hz), 1.94 (m, 2H), 1.28–1.38 (m, 16H), 0.94 (t,
3
3
6H, J=7.5 Hz), 0.88 ppm (t, 6H, J=6.3 Hz); 13C NMR (125.8 MHz,
CDCl3, 208C): d=160.3, 135.40, 126.5, 125.8, 46.3, 37.8, 30.6, 28.5,
23.9, 23.1, 14.1, 10.6 ppm; ATR-FTIR: n˜ =2959, 2924, 2910, 2855,
1709, 1661, 1585, 1458, 1435, 1406, 1373, 1308, 1283, 1192, 1155,
1139, 721, 582, 463 cmÀ1
.
1
vacuo (50 mg). Yield (C60H62B4F16N8NiO8P4): 85%. H NMR (300 MHz,
Synthesis of Dye-2: To a glass 20 mL microwave reactor 4,5,9,10-
tetrabromonaphthalene di(2-ethylhexyl)imide (130 mg, 0.16 mmol,
1.0 equiv.) and 3,4-diaminobenzene hydroxamic acid (108 mg,
0.64 mmol, 4.0 equiv) were added, together with degassed DMF
(10 mL). The yellow suspension was degassed for 5 min with
a stream of N2, and capped with a Teflon-lined stopper, after which
the vessel was heated using microwave irradiation for 30 min at
858C. The blue reaction mixture was cooled and poured into
CH2Cl2. The organic layer was extracted with saturated aqueous
NaHCO3, aqueous HCl (1m), and water. The organic layer was dried
over Na2SO4 and concentrated in vacuo. Column chromatography
(SiO2, 100% CH2Cl2 to 9:1 CH2Cl2/HOAc) afforded the product as
a dark blue solid. Yield 66.6 mg (51%). Poor solubility precluded
NMR spectroscopy of this material. M.p.:>3008C; IR: n˜ =2959,
2926, 2856, 1693, 1628, 1582, 1501, 1456, 1308, 1248, 1231, 1132,
CH3CN-d3, 258C): d=12.39 (br s), 7.83 (d, J=8.5 Hz, 8H), 7.55 (m,
6H), 7.30 (br m, 22H), 4.38 (d, J=14.3 Hz, 8H), 4.15 ppm (br d, J=
14.3 Hz, 8H); 13C NMR (300 MHz, [D3]CH3CN, 258C): 161.8, 154.9 (t,
J=4 Hz),132.2, 130.9, 130.0, 129.4, 124.5 (t, J=13 Hz), 113.5,
49.8 ppm (very br); 31P NMR (300 MHz, [D3]CH3CN, 258C): d=
À0.5 ppm (s); IR: n˜ =1602 (s), 1523, 1479 (s), 1436, 1417, 1393,
1200, 1056 (br s), 959, 878, 830, 738, 689, 518, 476 cmÀ1; HRMS
(ESI): m/z: calcd for [C60H59N8NiO8P4]+: 1201.2760; found:
1201.2846.
Synthesis of Dye-2
Synthesis of 3,4-diaminobenzene hydroxamic acid:[24] Methyl 3,4-
diaminobenzoate (1.66 g, 10 mmol) was dissolved in dioxane
(35 mL). Hydroxylamine hydrochloride (2.08 g, 30 mmol) and NaOH
(2.40 g, 10 equiv., 100 mmol) in degassed water (15 mL) was added
to this solution under stirring. After stirring at RT for 4 days, all vol-
atiles are removed in vacuo. The reaction mixture was neutralized
to a pH of approximately 7 through the cautious addition of con-
centrated HCl and stored at +48C to precipitate the product as
brown needles. The obtained solid was collected by filtration,
washed with a minimal volume of cold MeOH, and dried in vacuo.
1030 cmÀ1 HRMS (coldspray): m/z: calcd for [C37H42Br2N5O6]+:
;
812.14814; found: 812.14858; elemental analysis calcd for
C37H41Br2N5O6 (MW = 811.57): C 54.76, H 5.09, N 8.63; found: C
54.47, H 5.28, N 8.54.
Acknowledgements
This work is part of the ECHO research programme, which is
(partly) financed by the Netherlands Organisation for Scientific
Research (NWO). This project was carried out within the research
programme of BioSolar Cells, co-financed by the Dutch Ministry
of Economic Affairs. Z. Abiri is acknowledged for the synthetic ef-
forts contributing to this work. We thank Prof. Dr. A. M. Brouwer
and Dr. R. M. Williams for fruitful discussions, E. Zuidinga for
mass analysis, and L. de Lange for ICP analysis.
1
Yield 784 mg (47%). M.p.: 172–1778C (decomp); H NMR (500 MHz,
DMSO, 258C): d=10.64 (s, 1H), 8.56 (s, 1H), 6.97 (s, 1H), 6.85 (d,
3
3
1H, J=8.0 Hz), 6.44 (d, 1H, J=8.0 Hz), 4.94 (s, 2H), 4.56 ppm (s,
2H); 13C NMR (125.8 MHz, DMSO, 208C): d=165.8, 138.4, 134.0,
120.9, 116.6, 113.5, 112.8 ppm; IR: n˜ =3301, 3175, 1630, 1585, 1506,
1464, 1523, 1311, 1281, 1236, 1126, 1080, 1032, 892, 841, 810, 770,
731, 554, 515, 447 cmÀ1; HRMS (ESI): m/z: calcd for [C7H10N3O2]+:
168.0773; found: 168.0772.
Synthesis of 4,5,9,10-tetrabromonaphthalene di(2-ethylhexyl)i-
mide: 2-Ethylhexylamine (7.5 mL, 45 mmol) was added to de-
gassed glacial acetic acid (100 mL). 2,3,6,7-Tetrabromonaphtalene
dianhydride (8.4 g, 15 mmol) was added, and the mixture was
heated to reflux with a sand bath at 1508C. The initially turbid
yellow suspension turned into a clear, dark brown solution after
10 min of reflux, upon which the reaction vessel was cooled in an
ice bath. Immediate dilution with water (500 mL) water at RT pre-
cipitated a bright yellow solid, which was triturated with water,
aqueous NaHCO3, and cyclohexane on a Buchner filter. Vacuum
drying at RT gave a bright yellow solid (9.7 g), which could not be
analyzed satisfactorily by using NMR spectroscopy or MS. The
yellow solid was dissolved in degassed, distilled toluene (50 mL)
and heated to reflux. After 10 min, PBr3 (3.3 mL, 35 mmol) was
added dropwise with a syringe. 25 min after addition, the reaction
mixture was cooled and diluted under vigorous stirring with
NaHCO3 (500 mL) to quench excess PBr3. The aqueous phase was
Keywords: electrochemistry · dyes/pigments · hydrogen ·
nickel · photochemistry
[1] a) F. E. Osterloh, Chem. Soc. Rev. 2013, 42, 2294–2320; b) K. Maeda, K.
Domen, J. Phys. Chem. Lett. 2010, 1, 2655–2661; c) K. Sun, S. Shen, Y.
Liang, P. E. Burrows, S. S. Mao, D. Wang, Chem. Rev. 2014, 114, 8662–
8719; d) M. G. Walter, E. L. Warren, J. R. McKone, S. W. Boettcher, Q. Mi,
E. A. Santori, N. S. Lewis, Chem. Rev. 2010, 110, 6446–6473.
[2] For a recent review, see: Z. Yu, F. Li, L. Sun, Energy Environ. Sci. 2015, 8,
760–775.
[3] a) M. A. Gross, A. Reynal, J. R. Durrant, E. Reisner, J. Am. Chem. Soc.
2014, 136, 356–366; b) F. Lakadamyali, A. Reynal, M. Kato, J. R. Durrant,
E. Reisner, Chem. Eur. J. 2012, 18, 15464–15475; c) J. M. Gardner, M.
Beyler, M. Karnahl, S. Tschierlei, S. Ott, L. Hammarstrçm, J. Am. Chem.
Soc. 2012, 134, 19322–19325; d) E. Reisner, J. C. Fontecilla-Camps, F. A.
Armstrong, Chem. Commun. 2009, 550–552.
[4] a) L. Tong, A. Iwase, A. Nattestad, U. Bach, M. Weidelener, G. Gçtz, A.
Mishra, P. Bäuerle, R. Amal, G. G. Wallace, A. J. Mozer, Energy Environ. Sci.
ChemCatChem 2016, 8, 1392 – 1398
www.chemcatchem.org
1397
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
2012, 5, 9472–9475; b) C. E. Castillo, M. Gennari, T. Stoll, J. Fortage, A.
Deronzier, M.-N. Collomb, M. Sandroni, F. LØgalitØ, E. Blart, Y. Pellegrin,
C. Delacote, M. Boujtita, F. Odobel, P. Rannou, S. Sadki, J. Phys. Chem. C
2015, 119, 5806–5818.
[16] U. J. Kilgore, J. A. Roberts, D. H. Pool, A. M. Appel, M. P. Stewart, M. R.
DuBois, W. G. Dougherty, W. S. Kassel, R. M. Bullock, D. L. DuBois, J. Am.
Chem. Soc. 2011, 133, 5861–5872.
[17] A. Jain, S. Lense, J. C. Linehan, S. Raugei, H. Cho, D. L. DuBois, W. J.
Shaw, Inorg. Chem. 2011, 50, 4073–4085.
[5] a) M. L. Helm, M. P. Stewart, R. M. Bullock, M. R. DuBois, D. L. DuBois, Sci-
ence 2011, 333, 863–866; b) A. Dutta, D. L. DuBois, J. A. S. Roberts, W. J.
Shaw, Proc. Natl. Acad. Sci. USA 2014, 111, 16286–16291.
[18] As a result of very poor solubility, Ni-1 could not be studied by electro-
chemistry dissolved in water.
[6] L. Li, L. Duan, F. Wen, C. Li, M. Wang, A. Hagfeldt, L. Sun, Chem.
Commun. 2012, 48, 988–990.
[7] Z. Ji, M. He, Z. Huang, U. Ozkan, Y. Wu, J. Am. Chem. Soc. 2013, 135,
11696–11699.
[8] a) K. Fan, F. Li, L. Wang, Q. Daniel, E. Gabrielsson, L. Sun, Phys. Chem.
Chem. Phys. 2014, 16, 25234–25240; b) F. Li, K. Fan, B. Xu, E. Gabriels-
son, Q. Daniel, L. Li, L. Sun, J. Am. Chem. Soc. 2015, 137, 9153–9159.
[9] a) F. Odobel, L. Le Pleux, Y. Pellegrin, E. Blart, Acc. Chem. Res. 2010, 43,
1063–1071; b) F. Odobel, Y. Pellegrin, E. A. Gibson, A. Hagfeldt, A. L.
Smeigh, L. Hammarstrçm, Coord. Chem. Rev. 2012, 256, 2414–2423.
[10] G. Boschloo, A. Hagfeldt, J. Phys. Chem. B 2001, 105, 3039–3044.
[11] Q. Ye, J. Chang, K. Huang, X. Shi, J. Wu, C. Chi, Org. Lett. 2013, 15, 1194–
1197.
[19] J. A. Rombouts, J. Ravensbergen, R. N. Frese, J. T. M. Kennis, A. W. Ehlers,
J. C. Slootweg, E. Ruijter, K. Lammertsma, R. V. A. Orru, Chem. Eur. J.
2014, 20, 10285–10291.
[20] a) C. Haines, M. Chen, K. P. Ghiggino, Sol. Energy Mater. Sol. Cells 2012,
105, 287–292; b) S. Asir, A. S. Demir, H. Icil, Dyes Pigm. 2010, 84, 1–13;
c) C. Hippius, I. H. M. van Stokkum, E. Zangrando, R. M. Williams, M.
Wykes, D. Beljonne, F. Würthner, J. Phys. Chem. C 2008, 112, 14626–
14638; d) D. Veldman, S. M. A. Chopin, S. C. J. Meskers, M. M. Groene-
veld, R. M. Williams, R. A. J. Janssen, J. Phys. Chem. B 2008, 112, 5846–
5857; e) W. E. Ford, P. V. Kamat, J. Phys. Chem. 1987, 91, 6373–6380.
[21] a) C. J. Wood, K. C. D. Robson, P. I. P. Elliott, C. P. Berlinguette, E. A.
Gibson, RSC Adv. 2014, 4, 5782–5791; b) J. He, H. Lindstrçm, A. Hag-
feldt, S. Lindquist, J Phys Chem. B. 1999, 103, 8940–8943; c) P. Qin, H.
Zhu, T. Edvinsson, G. Boschloo, A. Hagfeldt, L. Sun, J. Am. Chem. Soc.
2008, 130, 8570–8571.
[12] D. W. Wakerley, M. A. Gross, E. Reisner, Chem. Commun. 2014, 50,
15995–15998.
[13] D. L. Ashford, M. K. Gish, A. K. Vannucci, M. K. Brennaman, J. L. Temple-
ton, J. M. Papanikolas, T. J. Meyer, Chem. Rev. 2015, 115, 13006–13049.
[14] a) Y. Gao, X. Ding, J. Liu, L. Wang, Z. Lu, L. Li, L. Sun, J. Am. Chem. Soc.
2013, 135, 4219–4222; b) W. R. McNamara, R. C. Snoeberger, G. Li, J. M.
Schleicher, C. W. Cady, M. Poyatos, C. A. Schmuttenmaer, R. H. Crabtree,
G. W. Brudvig, V. S. Batista, J. Am. Chem. Soc. 2008, 130, 14329–14338;
c) W. J. Youngblood, S. A. Lee, Y. Kobayashi, E. A. Hernandez-Pagan, P. G.
Hoertz, T. A. Moore, A. L. Moore, D. Gust, T. E. Mallouk, J. Am. Chem. Soc.
2009, 131, 926–927; d) W. R. McNamara, R. C. Snoeberger III, G. Li, C.
Richter, L. J. Allen, R. L. Milot, C. A. Schmuttenmaer, R. H. Crabtree, G. W.
Brudvig, V. S. Batista, Energy Environ. Sci. 2010, 3, 917–923.
[15] Y. Gao, L. Zhang, X. Ding, L. Sun, Phys. Chem. Chem. Phys. 2014, 16,
12008–12013.
[22] The Ni-1@NiO electrode has a very small emission compared to Dye-
2@NiO, and the blank NiO electrode practically does not emit under
the applied conditions.
[23] L. D’Amario, L. J. Antila, B. P. Rimgard, G. Boschloo, L. Hammarstrçm, J.
Phys. Chem. Lett. 2015, 6, 779–783.
[24] O. Kreye, S. Wald, M. A. R. Meier, Adv. Synth. Catal. 2013, 355, 81–86.
[25] X. Gao, C. Di, Y. Hu, X. Yang, H. Fan, F. Zhang, Y. Liu, H. Li, D. Zhu, J. Am.
Chem. Soc. 2010, 132, 3697–3699.
Received: January 8, 2016
Published online on February 23, 2016
ChemCatChem 2016, 8, 1392 – 1398
www.chemcatchem.org
1398
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim