Core-substituted naphthalenediimides anchored on BiVO4 for visible light-driven water splitting

Article abstract of DOI:10.1039/c7gc00125h

In this work, a novel catalytic system for the sunlight-driven water splitting reaction, which exploits the photocatalytic ability of BiVO₄ coupled to a new kind of noble-metal-free organic dye molecules, is proposed. Hence, mono- and di-substituted naphthalenediimides (NDIs) were designed to have different functional groups that provide to them both tunable optical properties and adjustable HOMO/LUMO levels, and were selectively prepared (starting from 1,4,5,8-naphthalenetetracarboxylic acid) achieving yields >69%. Smart anchoring groups (i.e. carboxylates or aromatic amines) were added to the dyes in order to allow them to covalently bond to acidic -OH groups present on the BiVO₄ surface. An easy and low-cost room temperature dip-coating technique was used to dye-sensitize both BiVO₄ powders and thin films. NMR, MS, FT-IR, TG, FESEM, XRD, XPS and optical analyses confirmed the successful organic synthetic routes and good dyes/BiVO₄ linkages. Photochemical and photoelectrochemical water oxidation reaction tests, together with DFT calculations, demonstrated that a proper alignment of the semiconductor/NDI-based dye energy levels is fundamental for enhancing the photocatalyst performance through a Z-scheme mechanism. The ability of the NDI organic molecules to delocalize the electronic charges was also a key factor for minimizing recombination processes and achieving more than a ten-fold increase in the photocurrent density of a 6 cm² BiVO₄ photo-electrode. The here reported results open new perspectives for the utilization of this new series of core-substituted NDIs, which are able to improve the activity of photocatalysts for different sunlight-driven applications, e.g. waste water treatment and organic contaminants' degradation, other than the production of solar fuels by water splitting mechanisms.

Full text of DOI:10.1039/c7gc00125h

View Article Online
View Journal
Green
Chemistry
Accepted Manuscript
This article can be cited before page numbers have been issued, to do this please use: S. Hernandez, C.
Ottone, S. Proto, K. Tolod, M. Díaz de los Bernardos, A. Solé-Daura, J. J. Carbó, C. Godard, S. Castillón, N.
Russo, G. Saracco and C. Claver, Green Chem., 2017, DOI: 10.1039/C7GC00125H.
This is an Accepted Manuscript, which has been through the
Volume 18 Number
7 7 April 2016 Pages 1821–2242
Royal Society of Chemistry peer review process and has been
accepted for publication.
Green
Chemistry
Accepted Manuscripts are published online shortly after
Cutting-edge research for a greener sustainable future
www.rsc.org/greenchem
acceptance, before technical editing, formatting and proof reading.
Using this free service, authors can make their results available
to the community, in citable form, before we publish the edited
article. We will replace this Accepted Manuscript with the edited
and formatted Advance Article as soon as it is available.
You can find more information about Accepted Manuscripts in the
author guidelines.
Please note that technical editing may introduce minor changes
to the text and/or graphics, which may alter content. The journal’s
standard Terms & Conditions and the ethical guidelines, outlined
in our author and reviewer resource centre, still apply. In no
event shall the Royal Society of Chemistry be held responsible
for any errors or omissions in this Accepted Manuscript or any
consequences arising from the use of any information it contains.
ISSN 1463-9262
CRITICAL REVIEW
G. Chatel et al.
Heterogeneous catalytic oxidation for lignin valorization into valuable
chemicals: what results? What limitations? What trends?
rsc.li/green-chem

Page 1 of 32
Green Chemistry
View Article Online
DOI: 10.1039/C7GC00125H
New Core-Substituted Naphtalenediimides anchored on BiVO4 for
Visible Light-Driven Water Splitting
Simelys Hernández,a,b,* Carminna Ottone,c Stefano Proto,d Kristhine Tolod,a
Miriam Díaz de los Bernardos,d Albert Solé-Daura,e Jorge J. Carbó,e Cyril
Godard,e Sergio Castillón,f Nunzio Russo,a Guido Saracco,b Carmen Claver,d,e*
a CREST group, Department of Applied Science and Technology (DISAT), Politecnico di Torino,
Corso Duca degli Abruzzi 24, 10129 Turin, Italy.
b
Center for Sustainable Future Technologies (CSFT@Polito), Italian Institute of Technology,
Corso Trento 21, 10129, Turin, Italy.
c School of Biochemical Engineering, Pontificia Universidad Catolica de Valparaiso. Av. Brasil
2085, Valparaiso, Chile.
d Centre Tecnològic de la Química a Catalunya (CTQC), CarrerMarcel·lí Domingo s/n, Edifici
N5, 43007, Tarragona, Spain.
e Departament of Química Física i Inorgànica, Universitat Rovira i Virgili, CarrerMarcel·lí
Domingo s/n, 43007, Tarragona, Spain.
f
Department of Química Analítica i Química Orgànica, Universitat Rovira i Virgili,
CarrerMarcel·lí Domingo s/n, 43007, Tarragona, Spain.
*corresponding authors: E-mail: simelys.hernandez@polito.it, carmen.claver@ctqc.org
Abstract
In this work, a novel catalytic system for the sunlight-driven water splitting, which exploits the
photocatalytic ability of BiVO4 coupled to a new kind of noble-metal-free organic dye molecules,
is proposed. Hence, mono- and di-substituted naphtalenediimides (NDIs), were designed to have
different functional groups that provide to them both tunable optical properties and adjustable
HOMO/LUMO levels, and were selectively prepared (starting from 1,4,5,8-
naphthalenetetracarboxylic acid) achieving yields > 69 %. Smart anchoring groups (i.e.
carboxylates or aromatic amines) were added to the dyes in order to allow them to covalently
bond to acidic -OH groups present on the BiVO4 surface. An easy and low-cost room temperature
dip-coating technique was used to dye-sensitize both BiVO4 powders and thin films. NMR, MS,
FT-IR, TG, FESEM, XRD, XPS and optical analyses confirmed the successful organic synthetic
routes and good dyes/BiVO4 linkages. Photochemical and photoelectrochemical water oxidation
reaction tests, together with DFT calculations, demonstrated that a proper alignment of the
semiconductor/NDI-based dye energy levels is fundamental for enhancing the photocatalyst
performance through a Z-scheme mechanism. The ability of the NDIs organic molecules to
delocalize the electronic charges was also a key factor for minimizing recombination processes
and achieving more than a ten-fold increase in the photocurrent density of a 6 cm2 BiVO4 photo-
1

Green Chemistry
Page 2 of 32
View Article Online
DOI: 10.1039/C7GC00125H
electrode. The here reported results opens new perspectives for the utilization of this new series
of core-substituted NDIs, which are able to improve photocatalysts activity for different sunlight-
driven applications, e.g. waste water treatment and organic contaminants degradation, other than
the production of solar fuels by water splitting mechanisms.
Keywords: Naphtalenediimides, BiVO4, photocatalysis, water splitting, DFT, dye-sensitized
semiconductor, EIS; Z-scheme.
1. Introduction
Light driven water-splitting into H2 and O2 is a topic of current interest,1-3 and the development
of a photocatalyst that can utilize the entire electromagnetic spectrum, in order to enhance the
overall efficiency of the process, is nowadays a challenging problem.4 In this context, bismuth
vanadate (BiVO4) has recently gained increasing attention for its use as a photo-anode capable of
promoting the oxidation of water to O2 in the half reaction of a complete PEC cell.5-8 BiVO4 is
one of the most promising photocatalyst for the oxygen evolution reaction (OER) because of its
relatively low band gap of 2.4 eV 9 for the monoclinic phase, which enables a more efficient use
of visible light, and due to the adequate position of the conduction and valence bands compared
to the potential of water oxidation. Moreover, BiVO4 is nontoxic and is composed of relatively
abundant elements. Regardless of the above quoted advantages provided by this inorganic
semiconductor, bare BiVO4 powders and photo-anodes have usually proved relatively poor
performances for photochemical and photoelectrochemical (PEC) O2 evolution due to slow O2
evolution kinetics, slow electron transport in the conduction band and, additionally, a significant
electron-hole recombination at both bulk and surface.10
Several attempts have been performed in order to obtain optimum crystal structures and
11-15
morphologies that are capable of enhancing the oxygen evolution activity of BiVO4.
Morphology control (heterojunction, homojunction, nanocasting, etc.)16-18 and doping
techniques19-22 have been investigated with the aim to enhance the BiVO4 photocatalytic
efficiency by improving both surface area and electron/hole mobility, which are necessary to
minimize recombination loses and facilitate the superficial charge transfer.
An alternative of this is the sensitization of the photocatalysts in order to use part of the solar
spectrum that would not be exploited otherwise. In order to warrant a good electron/hole transfer
between a sensitizer and the photocatalyst, the lowest unoccupied molecular orbital (LUMO) and
the highest occupied molecular orbital (HOMO) of the dye must have higher energy than the
conduction band (CB) and valence band (VB) of the semiconductor, respectively. Regarding the
last approach, several configurations for noble-metal-containing dyes have been reported (e.g.
2

Page 3 of 32
Green Chemistry
Ru, Rh- or Pd containing dyes).23, 24 However, few metal-free dyes have been reported as efficient
View Article Online
DOI: 10.1039/C7GC00125H
sensitizers for water splitting.25, 26 Besides, surface functionalization could represent a more
powerful and, at the same time, easily tunable approach able to improve the activity of BiVO4.
Indeed, recent developments in this context deals with the use of functionalized carbon dots (CDs)
for sensitizing semiconductor photo-electrodes.27, 28 Recently, the photocatalytic enhancement of
semiconductor based (i.e. TiO2) materials impregnated with a visible active polymer, i.e.
polyaniline have also been reported.29 Nevertheless, investigations concerning the use of organic
visible-light active species to provide dye-sensitized systems have received less attention so far.
Although Ru complex photosensitizers have shown very impressive conversion efficiency, their
large-scale application is limited because of costs and environmental issues. In this context, metal-
free organic dyes present many advantages, such as large molar extinction coefficient, control of
absorption wavelength, facile design and synthesis, and lower cost that make them competitive
with noble-metal-based complexes, in the near future.
In this work, we report our efforts toward the preparation of a novel catalytic system for light-
driven OER, which combines the proven ability of the inorganic BiVO4 in photochemical
transformations with an organic visible-light active species, i.e. core substituted
naphtalenediimides (NDIs). Four mono- and di-substituted NDIs were selectively prepared with
1,4,5,8-naphthalenetetracarboxylic acid as precursor. An appropriate control of the reaction
conditions, including bromination, condensation and bromide substitution, were performed in
order to obtain four metal-free dyes having different functional groups, which provide them
tunable optical properties (for an efficient light-harvesting of all the solar spectra) and adjustable
HOMO/LUMO levels. BiVO4, both powder and films, were impregnated with the dyes in order
to investigate the effect of the dyes on the photocatalytic responses. BiVO4 powder and photo-
30
6
anodes were prepared by previously optimized hydrothermal and spin-coating techniques,
respectively. The photocatalytic activity of BiVO4 impregnated with four different NDIs was
proven by evaluating both their photocatalytic and photoelectrochemical behavior for the OER
from water splitting under simulated sunlight irradiation. Their physico-chemical properties of all
the prepared heterostructures were fully characterized, as well.
2. Experimental Section
2.1. Syntheses of core-substituted naphtalenediimides (NDIs)
2.1.1. General.
Solvents and reagents were obtained from commercial sources (Sigma-Aldrich and Alfa-Aesar),
unless otherwise stated, and were purified and dried according to standard procedures.31 Air
3

Green Chemistry
Page 4 of 32
View Article Online
DOI: 10.1039/C7GC00125H
sensitive reactions were carried out under argon in dry solvents and glassware. Flash column
chromatography was carried out using a forced flow of the indicated solvent on Merck silica gel
60 (230−400 mesh). Reactions were monitored by TLC carried out on 0.25 mm E. Merck silica
gel 60 F254 glass or aluminum plates. Multinuclear NMR spectra were recorded on a Varian
1
Mercury 400 MHz instrument. Chemical shift values for H and 13C were referred to internal
SiMe4 (0.0 ppm). Significant peaks are tabulated in the order: multiplicity (s, singlet; d, doublet;
t, triplet; q, quartet; m, multiplet; br s, broad singlet), coupling constant(s) in Hertz (Hz) and
number of protons. High resolution electrospray mass spectra (HR-ESI) were collected in positive
mode on a MAT95XL from Thermo Finnigan. Matrix assisted laser desorption mass spectrometry
analyses were performed on a BrukerBiflex III MALDI-TOF spectrometer using dithranol as
matrix. The instrument was equipped with a nitrogen laser emitting at 337 nm, a 2 GHz sampling
rate digitizer, a pulsed ion extraction source and a reflectron. Fourier transform infrared (FT-IR)
spectra were recorded in pressed ATR mode on a Bruker Vector 22 spectrometer. TGA
experiments were performed by using the furnace of a Mettler Toledo TGA/SDTA851
instrument. X-ray powder diffraction (XRD) spectra were recorded using a Siemens EM-
10110BU D5000 diffractometer.
The synthetic strategy for the preparation of substituted-naphtalenediimides (dyes 1 to 4) is
schematized in the Fig. 1 and explained in the following sections 2.1.2 to 2.1.7.
2.1.2. Dye 1: N,N′-dipyridyl 1,4,5,8-naphthalenetetracarboxylic acid dianhydride
As shown in Fig. 1 (i), a mixture of 1,4,5,8-naphthalenetetracarboxylic acid dianhydride (0.8 g, 3
mmol) and 4-aminopyridine (0.56 g, 6 mmol) in anhydrous DMF (20 mL) was heated under reflux
for 8 h. When the reaction mixture reached room temperature (r.t.), a crystalline solid precipitated
out, which was collected by filtration. The crude product was purified by several recrystallization
from DMF to obtain the pure compound named dye 1 as pale yellow-orange needles (876 mg,
69% of yield).
2.1.3. 2,6-Dibromo-1,4,5,8-naphthalenetetracarboxylic acid dianhydride (Br,Br-NDA)
As shown in Fig. 1 (ii), a solution of dibromoisocyanuric acid (2.86 g, 10.0 mmol) in oleum (20
% SO3, 50 mL) was added at room temperature to a solution of naphthalene dianhydride (2.68 g,
10.0 mmol) inoleum (20 % SO3, 100 mL) over a period of 4 h. The resulting mixture was stirred
at r.t. for 1 h, then cautiously poured onto crushed ice (500 g) to give a bright yellow precipitate.
Water (1.5 L) was added, and the mixture was allowed to stand overnight at r.t. The yellow solid
was collected on a Büchner funnel and dried at air providing the crude product, which was used
for the next step without further purification. As it was not possible to purify at this stage, we
assumed quantitative formation. Mp: >350 °C. Due to its extremely low solubility, no NMR
spectrum could be obtained. It is worth to mention that batch-to-batch differences in the ratio of
4

Page 5 of 32
Green Chemistry
View Article Online
DOI: 10.1039/C7GC00125H
mono-, di- and tetra-Br NDAs were noticed, which seemed to reflect the sensitivity of this reaction
with respect to the concentration of the SO3 in oleum.
2.1.4. N,N’-Di-(2,6-diisopropylphenyl)-2,6-dibromonaphthalene-1,4,5,8-tetracarboxylic
bisimide (Br,Br-NDI)
acid
As shown in Fig. 1 (iii), a mixture containing 2,6-Dibromonaphthalene-1,4,5,8-tetracarboxylic
acid dianhydride (601.6 mg, 1.41 mmol) was suspended in acetic acid (30 mL) and heated at
reflux for 30 min. Freshly distilled 2,6-Diisopropylaniline (1.6 mL, 8.47 mmol) was added
dropwise and the mixture was kept at reflux at 120 ºC for 30 min. The solution was poured into
a mixture of crushed ice and distilled water and allowed to stand for 2 h at r.t. The red solid was
collected on a Büchner funnel, dried at the 60 ºC for 2 h and then purified on silica gel column
chromatography(hexanes:Et2O98:2) to give the desired compound as ared-orange solid (205 mg,
20% yield).
2.1.5. Dye 2: N,N’-Di-(2,6-diisopropylphenyl)-2-aminobenzoic acid-6-bromonaphthalene-
1,4,5,8-tetracarboxylic acid bisimide (Br,N-NDI)
For reaction (iv)(Fig. 1), in a 25 mL Schlenk-flask purged with Argon a mixture ofN,N’-Di-(2,6-
diisopropylphenyl)-2,6-dibromonaphthalene-1,4,5,8-tetracarboxylic acid bisimide (Br,Br-NDI)
(71 mg. 0.095 mmol) and 4-aminobenzoic acid (131 mg, 0.956mmol) in 3 mL anhydrous DMF
was stirred at 135 ºC. The clear red solution turned deep violet after 2 h and was cooled to room
temperature before quenching with aqueous saturated NaHCO3. The aqueous layer was extracted
with DCM (3 x 100 mL) and the combined organic extracts were dried over MgSO4, filtered and
concentrated under reduced pressure. The resulting waxy material was placed under high vacuum
until DMF was completely removed and a violet-blue solid was obtained. Purification by column
chromatography on silica gel (DCM:MeOH 9:1) provided dye 2 as a violet solid (71 mg, 93 %).
2.1.6. Dye
3:
N,N’-Di-(2,6-diisopropylphenyl)-2-aminobenzoic
acid-6-dimethylamine-
naphthalene-1,4,5,8-tetracarboxylic acid bisimide (N,N-NDI)
For reaction (v) (Fig. 1), in a 25 mL Schlenk-flask purged with Argon a mixture of N,N’-Di-(2,6-
diisopropylphenyl)-2,6-dibromonaphthalene-1,4,5,8-tetracarboxylic acid bisimide (Br,Br-NDI)
(40 mg. 0.054 mmol) and 4-aminobenzoic acid (295 mg, 2.15mmol) in 2 mL of anhydrous DMF
was stirred at 135 ºC. The clear red solution turned deep blue after 5h and was cooled to r.t. before
quenching with aqueous saturated NaHCO3. The aqueous layer was extracted with DCM (3 x 100
mL) and the combined organic extracts were dried over MgSO4, filtered and concentrated under
reduced pressure. The resulting waxy material was placed under high vacuum until DMF was
completely removed and a violet-blue solid was obtained. Purification by column
chromatography on silica gel (DCM: MeOH 9:1) provided dye 3 as a blue solid (40 mg, 96%).
5

Green Chemistry
Page 6 of 32
View Article Online
DOI: 10.1039/C7GC00125H
2.1.7. Dye 4: N,N’-Di-(2,6-diisopropylphenyl)-2-aminobenzoic acid-6-bromonaphthalene-
1,4,5,8-tetracarboxylic acid bisimide (Br,N-NDI)
For reaction (vi) (Fig. 1), in a 25 mL Schlenk-flask purged with Argon a mixture of N, N’-Di-
(2,6-diisopropylphenyl)-2-aminobenzoic acid-6-bromonaphthalene-1,4,5,8-tetracarboxylic acid
bisimide (Br,N-NDI) (75 mg. 0.094 mmol) and 4-aminobenzoic acid (129.2 mg, 0.94 mmol) in 4
mL anhydrous NMP was stirred at 150 ºC. The clear violet solution turned deep blue after 48h
and was cooled to r.t. before quenching with aqueous saturated NaHCO3. The aqueous layer was
extracted with DCM (3 x 100 mL) and the combined organic extracts were dried over MgSO4,
filtered and concentrated under reduced pressure. The resulting waxy material was placed under
high vacuum until NMP was completely removed and a blue solid was obtained. Purification by
column chromatography on silica gel (DCM:MeOH 9:1) provided dye 4 as a blue solid (73 mg,
90%).
2.2. Preparation of BiVO4 photocatalyst
2.2.1. BiVO4 powders
BiVO4 powders were prepared following a previously optimized receipt.30 Bismuth nitrate
pentahydrate (Bi(NO3)3·5H2O) and ammonium metavanadate (NH4VO3) both from Sigma
Aldrich were used as BiVO4 precursors. No further purification of the reagents were performed.
In a first step, 3.7 mmol of Bi(NO3)3·5H2O, 3.7 mmol of NH4VO3 and 12 mmol of ammonium
carbonate were dissolved in 75 mL of 1M HNO3. After 30 min of stirring at room temperature, a
clear solution was obtained (pH = 0). The mixture was then heated for 12h at 180 °C in a 100 mL
Teflon-lined stainless steel auto-clave under autogenous pressure. The precipitate was filtered,
washed three times with distilled water and then with ethanol, and dried at r.t. The dry samples
were annealed at 450 ºC for 2 h. All the chemicals used in the experiments were also of an
analytical grade and bi-distilled water was used for the preparation of the solution.
2.2.2. BiVO4 films
BiVO4 films were prepared by a spin-coating method into FTO-substrates over a 2 x 3 cm2 area.
The FTO-glass substrates were cleaned in acetone and ethanol by means of ultrasonication, and
then washed with bi-distilled water before the BiVO4 deposition. For a typical deposition, two
different solutions were prepared by dissolving 2.97 g of Bi(NO3)3·5H2O in 15 mL of acetic acid
and 1.624 g of vanadyl acetylacetonate (VO(acac)2) in 75 mL of acetylacetone. Such amounts are
doubled with respect to a recently published receipt,6 in order to increase the thickness of the film
by each single spin-coating step. The mixture of those solutions, which has a Bi/V atomic ratio
of 1, was subjected to ultrasonication for 30 min at 59 Hz to obtain a stable sol that was then spin-
coated at 500 rpm for 10 s into the FTO-glass substrates. After drying at 150 °C for 5 min on a
hot plate, the films were annealed at 400 °C for 10 min in order to induce the BiVO4
6

Page 7 of 32
Green Chemistry
View Article Online
DOI: 10.1039/C7GC00125H
crystallization. After natural cooling at r.t., this procedure was repeated one more time to increase
the film thickness. The electrodes were finally annealed at 400 °C for 2 h.
2.3. Preparation of naphtalenediimides decorated BiVO4.
The anchorage of the dye molecules to both the BiVO4 powderand film (see Fig. 2) was
accomplished by modified dip-coating techniques based on previously reported procedures.32, 33
Different solvent systems and conditions (i.e. Method A and Method B shown in Table 1) were
used depending to the solubility of the organic substrates. In a typical reaction, the dye was
dissolved in dry solvent and added via canula over 0.10 g of finely ground BiVO4 powder
previously charged in a 100 mL round-bottom flask, equipped with a reflux condenser and
maintained under inert atmosphere. The resulting suspension was stirred overnight and the solid
material was separated by filtration on a fritted glass funnel, wash with 20 mL of the respective
hot solvent solution and dried at 100 ºC for 1 h.
Table 1: Experimental conditions for the sensitization of BiVO4 with different dyes
Dye Concentration
[mM ]
Stirring Temperature
[ºC ]
Method
Dye
Solvents
Toluene : DMF
(50:3 v/v)
1
50
10
120
90
A
B
2, 3 or 4
Ethanol
2.4. Physicochemical characterization of naphtalenediimides-BiVO4
Crystalline structure and morphology of pristine and impregnated BiVO4 powders were
characterized by X-ray diffraction (XRD with a X’Pert PRO diffractomer) analysis by Cu-Kα X-
ray tube, λ = 1.54 Å with 40 kV of accelerating voltage and Field Emission Scanning Electron
Microscopy (FESEM) images performed with a Zeiss Merlin microscope.
UV-Vis reflectance spectra (% R) were collected with a spectrophotometer (Cary 500 by Varian)
by using an integration sphere. The surface composition of the samples was characterized by X-
ray photo-electron-spectroscopy (XPS, PHI5000 VersaProbe). The XPS analyses were carried
out with a VG Escalab 200-C X-ray photo-electron-spectrometer, with a non-monochromatic Mg-
K source. A pass energy of 20 eV, a resolution of 1.1 eV, and a step of 0.2 eV were used for the
high-resolution spectra. The effects of sample charging were eliminated by referring the spectral
line shift to /a C1sbinding energy value of 284.6 eV.
2.5. Water oxidation activity of naphtalenediimides decorated BiVO4
2.5.1. Photocatalytic Oxygen evolution measurements
The catalytic activity was measured by means of the previously reported Clark cell system.22 In a
typical test, 10 mg BiVO4 were tested in 11 mL of a 50 mM AgNO3 solution. Dissolved O2 (DO)
7

Green Chemistry
Page 8 of 32
View Article Online
DOI: 10.1039/C7GC00125H
was measured with a Clark type electrode (InPro 6050/120, Mettler Toledo). Prior to each
measurement, O2 was removed by flowing N2 through the liquid phase with a positive pressure
of 1.05 atm. Then, the reactor was illuminated with simulated solar light (AM 1.5G, 100mW/cm2)
with a plasma lamp (LIFI STA-40, LUXIM). O2 production rate (RO2) was measured as the slope
of the initial portion of DO curves [dDO/dt]0 and the max value of oxygen (DO*) was consider
as the total oxygen production. A control without catalyst was performed in dark and under
illumination in order to determine the increment of DO due to atmospheric oxygen.
2.5.2. Photoelectrochemical measurements
Photoelectrochemical (PEC) measurements were run with a standard three-electrode setup in a
lab-made cell. An Ag/AgCl (3 M KCl) electrode and a Pt wire were used as reference electrode
and counter electrode, respectively. The measurements were recorded with a multichannel VSP
potentiostat/galvanostat (BioLogic), equipped with EC-Lab v. 10.1x software for data acquisition.
Potentials vs. the Ag/AgCl reference electrode were converted to the reversible hydrogen
electrode (RHE) scale by using the Nernst Equation: ERHE =EAg/AgCl + E0Ag/AgCl + 0.059 pH. ERHE
is the converted potential vs. RHE (e.g. VRHE), E0Ag/AgCl is the standard potential of Ag/AgCl (3
M KCl) at 25 ºC, i.e. 0.209 V, and EAg/AgCl is the experimental potential measured against the
Ag/AgCl reference electrode (e.g. VAg/AgCl). The catalytic activity of BiVO4films was tested at pH
= 7.0 in 0.1 M sodium phosphate buffer. Linear sweep voltammetry (LSV) curves were performed
in the dark and under continuous or chopped simulated solar light, in the range between 0.4 VRHE
and1.6 VRHE (sweep rate of 20 mV s-1). The intensity of the light was maintained at 100 mW cm-
2 by adjusting the distance between the light source (i.e. Newport 450 W Xe lamp equipped with
an AM 1.5G filter simulating sunlight and a water filter) and the PEC cell. Chrono-amperometry
(CA) measurements were performed at 1.23 VRHE over chopped (2 min dark and 2 min light)
visible light irradiation. The charge transfer resistance across the semiconductor-electrolyte
interface was evaluated by EIS measurements, conducted in the 0.1 Hz to 0.5 MHz frequency
range, with an alternating current (AC) amplitude of 25 mV at different applied potentials (i.e.
0.63, 0.93, 1.23 and 1.53 VRHE). In order to measure the Incident photon-to-electron conversion
efficiency (IPCE) spectra, the PEC cell was illuminated using a Newport Xe lamp (150 W) by
varying the incident light from 300 nm to 700 nm (step size: 10 nm) by means of a monochromator
(Cornestone 130 by Newport), at an applied potential of 1.23 VRHE. In this case, the electrode
illuminated area was 1 cm2and the light power density was about 1 mW cm−2, measured at 390
nm.
2.6. Experimental determination of the HOMO and LUMO energy levels
The HOMO and LUMO energy levels and the band gap (Eg) between them were determined via
cyclic voltammetry (CV) by using the onset potential of the first oxidation and reduction peak,
8

Page 9 of 32
Green Chemistry
View Article Online
DOI: 10.1039/C7GC00125H
respectively. The estimations can be done with the empirical relation, shown in equation 1 and 2,
including the ferrocene (Fc) oxidation value of 4.8 eV as external standard.34 E1/2(Fc) is equal to
0.4 V.35
�
1
2
�
ꢀꢁꢂꢃꢂ = − �ꢀꢂꢄ − ꢀ
)� + 4.8 ꢇꢈ
(1)
(2)
(
ꢅꢆ
�
1
2
�
)� + 4.8 ꢇꢈ
ꢀꢉꢊꢃꢂ = − �ꢀꢋꢌꢍ − ꢀ
(
ꢅꢆ
Where EOX and ERED are the onset potentials [VAg/AgCl] of oxidation and reduction, respectively.
The experiments were performed in a three electrodes systems consisting of a Pt wire as the
counter electrode, an Ag/AgCl as reference electrode and a glassy carbon working electrode.
Three consecutives CVs with a scan rate of 100 mV/min between -3 VAg/AgCl and 4 VAg/AgCl were
carried out. For the preparation of the sample, 20 µꢎ of a solution containing the dye (10 mg/ml)
was dropped on a glassy carbon electrode and left to evaporate in dark conditions. The supporting
electrolyte was a solution of 0.1 M tetrabutylammonium hexafluorophosphate (TBAPF6) in
acetonitrile. To ensure that there is no oxygen in the electrolyte, a purge with nitrogen was
performed for 15 min at a flow rate of 24 Nml/min before each measurement.
2.7. Computational Details
All calculations were performed using the Gaussian09 series of programs.36 Full quantum
mechanics calculations on the energy levels of the dyes were performed within the framework of
density functional theory (DFT) using the BP86 37, 38 and PBE 39, 40 functionals with the basis set
6-31g(d,p),41-43 for all the atoms. All geometry optimizations were full, with no restrictions using
the Berny algorithm implemented in Gaussian09.44 The solvent effects of the acetonitrile was
computed via the continuum IEF-PCM model.45
3. Results and Discussion
3.1 Synthesis of core-substituted NDIs.
The syntheses of the dyes 1 - 4 described in this work were performed according to the scheme
represented in Fig. 1. The condensation between the commercially available 1,4,5,8-
naphthalenetetracarboxylic acid dianhydride and 4-aminopyridine in anhydrous DMF, provided
the first target compound dye 1 with 69 % of yield after several recrystallization steps as
confirmed by NMR, FT-IR and MS analyses (see section 2.1.2).
The preparation of the key intermediate (i.e. 2,6-dibromo-naphthalene tetracarboxylic
dianhydride) for the syntheses of the dyes 2 to 4 was accomplished by the bromination of the
starting anhydride with dibromoisocyanuric acid (DBI) in oleum. A mixture of poly-brominated
products were obtained and were directly employed in the next condensation with 2,6-di-
isopropyl-aniline. To avoid SNAr reactions, the latter step was performed in acetic acid, where
9

Green Chemistry
Page 10 of 32
46
View Article Online
DOI: 10.1039/C7GC00125H
most amino functions were protonated. The reaction product was isolated with 54% of yield and
confirmed by NMR, F-TIR and MS analyses (see supporting information, SI).
Subsequently, the substitution of the two bromide substituents was achieved using various
reagents to produce the dyes 2 to 4 (see sections 2.1.5 and 2.1.6). The novel mono- and di-
substituted cNDIs (i.e. dyes 2-4) were selectively prepared from the same intermediate and 4-
aminobenzoic acid through an appropriate control of the reaction conditions (time and equivalents
of 4-aminobenzoic acid). It is worth to illustrate that the introduction of the NMe2 functionality
by SNAr substitution, occurred unexpectedly from DMF decomposition. Indeed, some literature
reports 47, 48 described that under specific reaction conditions DMF could decompose providing a
source of–NMe2 group, which in our case would be the responsible of the second substitution on
the naphthalene core.
The purity and composition of all the prepared NDIs were confirmed by NMR and FT-IR
spectroscopy as well as by MS characterization (Fig.s S1 to S5 in SI) and an excellent yield of the
syntheses was obtained: 93%, 96% and 96%, for dyes 2, 3 and 4, respectively.
Moreover, the new synthesized NDIs evidence light absorption over a broad range of wavelengths
in the UV and Vis ranges (see Fig. S6, SI); i.e. between 200 and 390 nm for all the dyes and with
peaks in the visible region for dyes 2 to 4. Indeed, the dye 1 evidence main adsorption peaks in
the UV range, at wavelengths (λ) of 237 and 350 nm, as expected for core-unsubstituted NDIs. 46
Instead, dyes 2, 3 and 4 present significant bathochromical shift of the absorption maxima,
attaining noticeable visible light harvesting peaks at 520, 600 and 660 nm, respectively, other
than in the UV region, because of the different functionalization introduced to the naphthalene
core. 46 Interestingly, the dye 4 practically cover all the solar spectra with a more constant light
absorption intensity. The here synthesized core-substituted NDIs are thus suitable to cover the
“green gap”, i.e. operate at wavelengths longer than 500 nm, hence, being promising materials to
be used into photocatalytic systems with other semiconductor materials.
The resulting optical properties of the dyes obviously depend on their HOMO and LUMO
energies and band gap (Eg), which also are close-correlated to the photocatalytic behavior of the
dye-sensitized BiVO4 photocatalyst and, therefore, will be introduced above in the section 3.3.
3.2 Structural, optical and chemical characterizations of NDI-decorated BiVO4.
The efficient anchoring of the NDI dyes onto the BiVO4 surface was confirmed by XRD, FT-IR,
TGA, FESEM and XPS analyses. The quantification of the amount of each dye in the
corresponding functionalized material was obtained by thermogravimetric analyses (see Fig.s S7b
and S8, SI). The highest impregnation, equal to 42 wt%, was obtained with dye 1, followed by
dye 4, dye 3 and dye 2 with a 14 wt%, 6 wt% and 3.8 wt%, respectively. The highest impregnation
obtained with dye 1 could be attributed to the highest concentration used during the impregnation,
which was 5-fold higher than the one used with dyes 2 - 4. Nevertheless, the different functional
10

Page 11 of 32
Green Chemistry
View Article Online
DOI: 10.1039/C7GC00125H
groups presents in the dyes 1 to 4 have also a direct effect on the binding interactions formed with
the BiVO4 surface and, consequently, on the final dye loading.
Previous FT-IR analysis and amines absorption test carried out on BiVO4 samples revealed the
presence of Lewis acid sites on its surface,49 that are supposed to be the main anchoring moiety
during the functionalization of the semiconductor. In the case of the simple core-unsubstituted
NDI (dye 1), which contains two pyridine moieties, it is possible that this kind of linker
established a stronger interaction with the BiVO4 surface by means of the two exposed aromatic
nitrogen atoms. Instead, in the case of the dyes 2 - 4, the carboxylic acid functionality, which is
the most commonly used anchoring group for dye sensitized solar cells,33 is envisaged to be the
linkage moiety with the BiVO4 surface.
From a structural and morphological point of view, the bare BiVO4 powder exhibits a monoclinic
scheelite crystalline phase (see FESEM image, Fig. 3a, and XRD spectrum, Fig. 4a). BiVO4
powder’s diffraction peaks were in good agreement with the standard Joint Committee on Powder
Diffraction Standards (JCPDS) card No. 14-0688 (space group: I2/a, a = 5.195, b = 11.701, c =
5.092, β = 90.38 o) and shows the characteristic peak splitting diffractions at 2θ = 18.5, 35 and 46
o.30 No traces of scheelite–tetragonal or zircon–tetragonal phases are observed in the diffraction
pattern, which points out the high selectivity of the chosen synthesis method. As shown in Fig. 3,
after the impregnation with the dyes, the level of coverage of the BiVO4 crystals varies with the
dye used. In particular, the surface of the BiVO4 crystals impregnated with the dye 1 is completely
covered by a needle-like crystalline structure (Fig. 3b). Coherently, the X-ray diffraction of the
o
BiVO4 + dye 1 powder shown in Fig. 4a evidences three peaks at 19.2, 22.8 and 23.6 that
correspond to the crystalline structure of dye 1, which XRD complete spectrum is exhibited in
Fig. S7c (SI). In contrast to dye 1, dyes 2 - 4, have an amorphous structure, forming an external
layer on the photocatalyst that fill the BiVO4 inter-particles spaces. The image corresponding to
the BiVO4 powder impregnated by the dye 4 is exhibited as an example in Fig. 3c, which is
representative of what was observed after the impregnation with dyes 2 and 3, as well. Fig. 3c
shows the BiVO4 crystals with grain boundaries that are less defined than such reported in Fig 3a
(for bare BiVO4) and are evidently embedded in the non-crystalline shell constituted by the dye.
Indeed, due the covering of the BiVO4 powder with the dyes 2 to 4 less defined BiVO4 XRD
peaks were observed (Fig. 4a), which again is in agreement with a good coverage of the BiVO4
crystals by these dyes.
The UV-Vis spectra, in reflectance percentage (% R), of the bare BiVO4 in comparison with the
dye-anchored powders are shown in Fig. 4b. These curves are inversely proportional to the
absorption spectrum. As expected and in agreement with previous results, the impregnation with
the dyes 2 - 4 increases the absorption region of BiVO4 to the visible region. This was not
11

Green Chemistry
Page 12 of 32
View Article Online
DOI: 10.1039/C7GC00125H
observed with dye 1, which mainly absorbs in the UV-region (see Fig. S6). Interestingly, the dye
3 maintains the absorption in the UV-region, whereas dyes 2 and 4 showed less absorption than
bare BiVO4 in the range of λ < 500 nm.
The surface coverage of the BiVO4 with the dye is confirmed by the XPS spectra as well (Fig. 5).
The intensities of Bi 4f y V 2p spectra, corresponding to elements present only in BiVO4 molecule,
decreases in the order: bare BiVO4 > BiVO4 + dye 3 > BiVO4 + dye 2 > BiVO4 + dye 1 > BiVO4
+ dye 4. From the C 1s spectra of the samples, three peaks are observed, where the peak at 284
eV is attributed to adventitious carbon contamination. In addition, the peaks, observed in the
positions 286 eV and 288 eV for C 1s and 532 eV and 533 eV for O 1s, are attributed to the
organic C-O and C=O bonds, respectively, and are present (as expected) only in the dye-
impregnated powders. Moreover, in the O 1s spectra (Fig. 5b), the peak at 530 eV represents the
binding energy of oxygen in metal oxides that, in this case and in line with the previous results,
shows the highest intensity for the bare BiVO4 sample. Finally, the highest intensity for the N 1s
high resolution XPS spectra, presented in Fig. 5c, was observed for dye 1, followed by dye 4, dye
3 and dye 2; interestingly, following the same trend observed in the weight percentage of the dyes
loaded into the BiVO4 powder measured by TGA.
3.3 Water oxidation photocatalytic performances of NDI decorated BiVO4 particles
The photocatalytic activity of the bare and dye-impregnated BiVO4 powders was measured by
means of a Clark type photo-reactor (see Fig. 6a), where atmospheric oxygen was stripped by a
constant nitrogen flow which kept a positive pressure (1.05 bar) inside the reactor. Dissolved
oxygen curves are shown in Fig. 6b, which were obtained in the curse of the experiment with a
Clark electrode. After 10 minutes under dark conditions, the reactor was illuminated with 1 sun
of irradiance and, in that instant, the curves show a sudden increment in the oxygen production.
It is worth noting that the slight increment of the oxygen in dark conditions is due to the baseline
of atmospheric O2. The oxygen evolution rate (RO2) and the maximum evolved oxygen (DO*) are
reported in Fig. 6c. RO2 were calculated as the slope of the first stage of the increasing section of
DO curves (10 min < t < 15 min), the values are plotted in Fig. 6c.
It is worth noting, that there is either no correlation between the DO* and the RO2, nor between
the catalytic performance and the dye loading into the BiVO4, as it has been observed in previous
works by using MnOx and [Ru(bpy)3]2+ for the OER, where a linearity between these parameters
was found.50 This is an indication of the fact that each dyes has a different interaction and effect
on the BiVO4 photocatalyst surface.
In addition, the experiments in the Clark reactor revealed that, although the RO2 of the BiVO4 does
not shown a particular improvement induced by any of the dyes, the BiVO4 + dye 3 sample
exhibits the maximum oxygen production. It was even 75 % higher than the bare BiVO4, whereas
the DO* obtained by the impregnation of the dyes 1, 2 and 4 remained lower than that of bare
12

Page 13 of 32
Green Chemistry
View Article Online
DOI: 10.1039/C7GC00125H
BiVO4. This result agrees with the UV-Vis spectrum of dye 3 (Fig. 4b), which showed that the
dye impregnated powder can absorb in the visible region but without reducing the absorbance in
the UV region, which could mean that the light harvesting of BiVO4 is not blocked by the dye. In
contrast to the BiVO4 samples containing the dyes 2 and 4, which also showed an increased UV
and Vis light absorption, respectively, but a poorer reacting performances, the dye 3 seems to
have a better electronic interactions with BiVO4, which is reflected by a higher total oxygen
production.
3.4 Energy levels of NDIs and water oxidation mechanism of BiVO4 + dye powder
photocatalyst
To better understand the interaction between BiVO4 and the different dyes, the energy levels of
HOMO and LUMO of dyes 1 - 4 were calculated and compared with the energy levels of BiVO4
and the water splitting reaction. The values of the energy levels were determined by two methods,
experimentally and theoretically, the latter by means of DFT calculations (see section 2.7).
Table 2 compares the energy of the HOMO and LUMO orbitals determined experimentally and
computationally through two different non-hybrid DFT functionals (BP86 and PBE) , which were
shown to be relatively accurate for predicting HOMO-LUMO gaps.51
Table 2: Values of HOMO and LUMO energies and band gap (Eg) for dyes 1 - 4 obtained by
both experimental (Exp.) and theoretical methods: BP86 and PBE functionals. See further details
in Table S1 and Fig. S9 (SI).
HOMO [eV]
Exp.
-6.35
-5.53
-5.16
LUMO [eV]
Exp.
-3.66
-3.95
-3.65
Eg [eV]
Exp.
2.69
1.58
1.51
BP86
-6.20
-5.49
-4.94
-5.09
PBE
-6.04
-5.39
-4.84
-4.98
BP86
-4.13
-4.01
-3.69
-3.88
PBE
-4.04
-3.90
-3.58
-3.78
BP86
2.07
1.48
1.26
1.21
Dye 1
Dye 2
Dye 3
Dye 4
-5.40
-3.82
1.57
The experimental and computed energies of the HOMO follow the same trend: dye 1 << dye 2
< dye 4 < dye 3. The computed absolute values are qualitatively, if not quantitatively, close the
experimental ones, mutually supporting both methodologies (see Table 2). The largest
discrepancy between the experimental and calculations appears for the LUMO energy of dye 1,
which is the lowest among the series of dyes according to DFT methods, following the same trend
as the HOMO energies, but it is one of the highest according to experimental method. Apart from
this case, the calculated LUMO levels are in rather good agreement with experimental values
obtained with experimental method. It is important to note that time-independent quantum
methods were found to give a poorer description of unoccupied than occupied orbitals.51
13

Green Chemistry
Page 14 of 32
View Article Online
DOI: 10.1039/C7GC00125H
Fig. 7 depicts the DFT computed frontier molecular orbitals and their respective position with
respect to BiVO4 and water splitting reaction by using the average value between the BP86 level
and experimental data (see Table S1, SI). Within the series of studied dyes, the addition of
electron-donating substituents to the naphthalene core in dyes 2 - 4 destabilizes significantly the
energy of the HOMO with respect to dye 1. For dyes 2 - 4, the HOMO is delocalized over the 4-
rings core forming π-antibonding interactions with the electron-donating substituents, whereas in
dye 1 the HOMO is delocalized over the two pyridine substituents. Roughly, the energy of the
HOMO (dye 1 << dye 2 < dye 4 < dye 3) correlates with the donor ability of the corresponding
dye substituents (H < Br < N(H)ArCOOH < NMe2). DFT calculations show the same
destabilization trend for the LUMO energies. However, the energy of the LUMO is less altered,
and consequently, the energy gaps for the dyes with electron-donating substituents (dyes 2 - 4)
are significantly narrower than for dye 1. This trend is also observed for experimental energy gap
values despite the above mentioned discrepancy on the LUMO energies for dye 1.
In order to establish if the dyes 1 - 4 work as sensitizer, some physical requests have to be
satisfied. The electron transfer from the donor to the acceptor requires that the lowest unoccupied
molecular orbital (LUMO) level of the dye is above that of the BiVO4 (i.e. its CB), and hole
transfer from the acceptor to donor requires that the highest occupied molecular orbital (HOMO)
level of the dye is above that of the BiVO4 (i.e. its VB), in order to fulfill the energy requirements
and thus to generate the photocurrent. From Fig. 7 we can inferred that dyes 1 - 4 fulfil those
requisites.
A Z-scheme type photosystem, which has demonstrated enhanced photocatalysts performances
by mimicking the natural photosynthesis in Photosystem II,52 is possible with the here presented
dyes and could explain why the WO performance of the BiVO4 +dye 3 couple is better than such
with the other dyes. As schematically depicted in Fig. 8a, the forward intramolecular and
interfacial hole/electron transfer processes can be summarized as follows: (1) upon light
irradiation the dye produces a long-lived charge separated excited state (dye + hν → dye*), which
eventually could evolve to the minimum on the triplet potential energy surface, and it results in a
photo-excited electron, dye(e-), and a hole, dye(h+); (2) the BiVO4 also absorbs radiation injecting
an electron in the conduction band, BiVO4 (e-), and generating a hole in the valence band, BiVO4
(h+), which is able to oxidize water (BiVO4 + hν → BiVO4*); (3) electron-hole recombination to
the excited dye in order to preserve the hole in the BiVO4, so that it can act as a water oxidation
catalyst (BiVO4* + dye(h+) → BiVO4(h+) + dye(e-)); (4) electron injection into the electron
acceptor (AgNO3 salt) or into the conduction band of the semiconductor oxide (dye(e-) + Ag+ →
dye + Ag0, or dye(e-) + BiVO4 → dye + BiVO4(e-)). The non-optimal performance of the overall
process can be attributed to the radiative and non-radiative recombination within the BiVO4 oxide
(3’ in Fig. 8a). However, the presence of the dye can hinder the electron-hole recombination
14

Page 15 of 32
Green Chemistry
View Article Online
DOI: 10.1039/C7GC00125H
within the BiVO4 oxide by accepting the photo-excited electron of BiVO4 in the hole produced in
the dye HOMO after its photo-excitation. This process is likely to be maximized when the
conduction band of the oxide and the HOMO of the dye are close in energy. Thus, since dye 3 has
the strongest e- donor substituent (NMe2) in the naphthalene core what destabilizes
significantly the energy of the HOMO, it shows the lowest differences between HOMO and CB
energy levels. Therefore, a better electron transfer with BiVO4 with respect to the other dyes is
allowed, which could explain the better performances obtained with the BiVO4 + dye 3 in the
Clark-reactor experiments.
From the Fig. 7 it can also be noticed that holes transfer from the BiVO4 to the dye 3 cannot cause
the water splitting reaction in the surface of the dye, because the energetic level of holes in the
HOMO of dye 3 are not enough oxidant to induce e- transfer from the water to that dye, with a
consequent impossibility to achieve water splitting and O2 evolution. In order to better clarify the
actual mechanism of charges transfer and the real possibility to implement the here developed
dyes in a photoelectrochemical device, the PEC behavior of the most promising dye 3 coupled
with a BiVO4 photo-electrode has been evaluated, and the results are presented in the next section
3.5.
3.5 Photoelectrochemical application of the NDI dye 3-anchored BiVO4 film
In order to assess the possible practical application and stability of the NDIs as well as to confirm
the capability of the best NDI dye to enhance the reaction kinetics, the performance of a BiVO4 +
dye 3 film was evaluated and compared to such of bare BiVO4 for the overall water splitting
reaction. The PEC cell containing a 0.1 M Na-phosphate buffer solution (pH 7). From both
chopped and continuous light LSV scans (Fig. 9a), a gradual increase of the photocurrent under
simulated sunlight irradiation (100 mW cm-2) is observed for both samples, with an onset potential
at about 0.6 VRHE. A reduction of the theoretical redox potential for water oxidation (Eº = 1.23
VRHE) indicates that part of the energy required for the reaction is provided by the simulated
sunlight.
The quite low photocurrent density (at 1.23 VRHE) of the BiVO4 film (∼ 0.03 mA·cm-2), is in
agreement with previous data in the literature for high area electrodes such the ones here
employed (2 x 3 cm2).6 Nonetheless, the implementation of the here developed dye 3 over a broad
electrode surface was consciously made to test a system closer to a practical case. As noticed from
Fig. 9, the generated photocurrent varies considerably between the bare and impregnated BiVO4
films. The BiVO4 + dye 3 film shows a photocurrent of 0.3 mA cm-2 at 1.23 VRHE, which
correspond to more than ten-fold the photocurrent generated with the bare BiVO4. In addition, the
photocurrent of the BiVO4 + dye 3 film does not arrive to a plateau, such as the bare BiVO4 does,
but continuous an increment by increasing the potential, reaching up to 0.53 mA cm-2 at 1.6 VRHE
.
15

Green Chemistry
Page 16 of 32
View Article Online
DOI: 10.1039/C7GC00125H
Fig. 9b shows the photocurrent density stability of the studied photo-anodes as a function of time.
To perform this, the samples were illuminated in cycles of two minutes of darkness and simulated
solar light irradiation. The photocurrent shows an instantaneous increment when the irradiation
begins, then shows a slight decrease and stabilizes to a steady state condition after few seconds.
The BiVO4 + dye 3 film shows a more pronounce spike respect to the bare BiVO4, showing a
slight lower photocurrent than the LSV curves. Such is commonly observed because in chopped
LSVs the irradiation time is too short for the sample to reach the steady state condition as well as
the continuous increase of the applied potential induces a faster e- transport in LSVs than in CAs.
However, the enhancement in the photocurrent due to the dye 3 impregnation is in agreement
with the LSV results. It is important to notice that, the photocurrent decreased quickly down to
zero when the light is turned off, confirming the effective photocatalytic (and not only catalytic)
activity of both studied films.
In order to identify the portion of the solar spectra that is actually working in the BiVO4 + dye 3
film in comparison to the bare BiVO4 film, the photo-response of the different photo-anodes have
been evaluated through IPCE spectra at an applied potential of 1.23 VRHE (Fig. 10a). IPCE curves
revealed that both the BiVO4 and BiVO4 + dye 3 have a relevant efficiency in the UV-Vis region
of sunlight up to a maximum of 500 nm. The maximum IPCE was obtained at 320 nm for both
samples and is of about 28 % and 49 % for the pristine BiVO4 and the dye 3-impregnated film,
respectively. Hence, there was no evidence of an increase of the action spectra of BiVO4 in the
visible range due to the presence of the dye, which is actually in agreement with the previous
discussion related to the position of the HOMO level of this dye and the inability to achieve water
oxidation by the holes on its surface. Nevertheless, the enhanced activity of the BiVO4 + dye 3
material is also confirmed by the IPCE measurements.
The better feature of the dye 3 impregnated film could be explained by the different transport
properties of such electrode, which can be observed from the Nyquist plots of the electrochemical
impedance spectroscopy measurements under simulated sunlight illumination reported in Fig.
10b. The smaller impedance values reported by the BiVO4 + dye 3 film than by the bare BiVO4,
at the different applied potentials, indicate a faster electron transfer (higher reaction kinetics at
the electrode/electrolyte interface) by reducing e-/h+ recombination’s on the BiVO4 surface.6, 53, 54
For instance, at 1.23 VRHE the charge transfer resistance of the BiVO4 + dye 3 film (∼ 136 Ω) is
less than a half of such of the bare BiVO4 (∼ 300 Ω).
The mechanism of improvement of charge transfer in the surface of the BiVO4 by the dye 3 can
be explained through two phenomena. Other than the previously explained Z-scheme that is
possible between the BiVO4 and dye 3, the dye 3 could also act as an enhancer of electrons
mobility and, consequently, of charges separation and photocurrent generation by a higher
16

Page 17 of 32
Green Chemistry
View Article Online
DOI: 10.1039/C7GC00125H
availability of holes in the BiVO4 surface able to react with water. The latter phenomena is similar
to such previously observed for BiVO4 films with reduced graphene oxide55 or in TiO2 porous
films impregnated with a conductive polymer (i.e. polyaniline).29 In the present case, the unique
characteristic provided by the aromatic diimide moiety (i.e. NDI) and the core-substituted
derivatives, which have been already exploited as electron-acceptor moieties in artificial light-
harvesting systems such as dye sensitized solar cells, is responsible for the semiconducting
property of the dye 3.56 Indeed, the planar geometry of NDIs and their low reduction potential
have been reported to provide a large driving force for the charge separation process, 56, 57 since
NDIs can be reversibly reduced to stable radical anions, under mild conditions.
The possible mechanism by which dye 3 improve the photo-electrocatalytic activity of BiVO4
under sunlight irradiation is schematized in Fig. 8b. When the BiVO4 + dye 3 heterostructure is
irradiated by UV-Vis light, the dye 3 absorbs photons and electrons in the HOMO can be excited
to the LUMO. In the same manner, BiVO4 absorbs UV-Vis photons and electrons in the BiVO4
VB can be excited to its CB, thereby leaving holes at its surface able to split water to O2. Then,
through the Z-scheme process, electrons in the CB of BiVO4 can be injected in the HOMO of dye
3, because they are at a higher energy level than the BiVO4 VB; otherwise, electrons in the dye 3
LUMO can be injected into the BiVO4 CB and then, due to the electromagnetic forces induced
by the applied bias, be injected into the FTO conductive substrate. These processes are beneficial
for an efficient charge carrier’s separation because they induce an increased electrons mobility
and transport in the BiVO4 CB. 29, 56 Hence, dye 3 likely hinders the recombination of the e-/h+
pairs and lengthen the lifetime of the charge carriers, consequently improving the charge
separation and photocatalytic activity of the BiVO4 photocatalyst. Thus, the fast transferred
electrons into the FTO conductive substrate are then conducted through the external electrical
circuit to the Pt cathode where the H2 evolution reaction takes place.
4. Conclusions
In this work, a new series of core-substituted NDIs was designed, with the aim of obtaining
tunable absorption properties that overlap with all the solar spectrum and bearing appropriate
functionalities to provide efficient and stable linkage on the surface of the BiVO4 water-splitting
photocatalyst. Four different molecules were successfully synthesized achieving excellent yields
(∼ 69 % for dye 1 absorbing UV-light; and > 93 % for dyes 2 to 4 absorbing visible light) as it
was confirmed by NMR, FT-IR and GC-MS analyses. The surface functionalization of BiVO4
has been accomplished by an easy and low-cost room temperature dip-coating technique. The
successful anchoring, as well as the stability of the photocatalyst-dye interactions, have been
successfully confirmed by TGA, FT-IR, XRD, XPS and optical analysis.
17

Green Chemistry
Page 18 of 32
View Article Online
DOI: 10.1039/C7GC00125H
The diversely sensitized BiVO4 samples were tested for the photocatalytic oxygen evolution
process under simulated sunlight conditions by employing two different systems: a Clark-type
reactor with a sacrificial e- acceptor (AgNO3) and a photoelectrochemical water splitting cell. The
optimum dye-sensitized catalytic systems was proven to be such including dye 3 on BiVO4, which
provided a remarkable enhancement in the O2 evolution from both quantitative and kinetic point
of views, confirming the effectiveness of the catalytic system here designed. The superior activity
4, 58, 59
of the BiVO4 + dye 3 system has been ascribed to two phenomena: first, a Z-scheme
mechanistic approach is possible due to the right position HOMO and LUMO energy levels of
the dye 3 with respect to those of the BiVO4 (i.e. CB and VB), as determined by DFT calculations
and experimental data; moreover, we considered that the ability of the here developed organic
dyes to delocalize the electronic charge and enhance electrons transport towards the FTO substrate
could be an additional key factor that minimizes recombination processes in the BiVO4 + dye 3
water oxidation photoanode. The here proposed approach opens new perspectives in the
investigation of a different series of dye molecules (namely, NDIs) that are low-cost, because do
not contain expensive noble metals, and are able to improve the activity of photocatalysts for
different sunlight driven applications, such as waste water treatment and organic contaminants
degradation, other than the production of solar fuels by water splitting mechanisms.
Acknowledgements
Sitaramanjaneya Mouli Thalluri, Mauro Raimondo and Salvatore Guastella are acknowledged for
the technical support during the synthesis of the BiVO4 powders, FESEM and XPS measurements,
respectively. This work has been funded by the EU Commission, 7th European Framework
Program (NMP-2012), Project Eco2CO2, Grant Agreement number 309701. Computational
research was supported by the Spanish Ministry of Science and Innovation (CTQ2014-52774-P)
and the Generalitat de Catalunya (2014SGR199 and XRQTC). The CTQC/URV authors
acknowledge the Spanish Ministerio de Economía y Competitividad (CTQ2013-43438-R) and
the Generalitat de Catalunya (2014SGR670) for financial support.
Supporting Information (SI)
Supplementary Information is available: NMR, FT-IR spectra and MS analyses of dye 1 to 4 and
Br,Br-NDI; photographs of dyes solutions in acetonitrile, their UV-Vis spectra and molecular
structures; FT-IR spectra and TGA of BiVO4 + dyes heterostructures; theoretical and
experimental data related to the calculation of the energy levels of the dyes. See DOI:
10.1039/x0xx00000x.
Author Contributions
18

Page 19 of 32
Green Chemistry
View Article Online
DOI: 10.1039/C7GC00125H
S. Hernández optimized the synthesis of BiVO4 powders and films with advises of N. Russo; she
also designed, coordinated and supervised the photo and electro catalytic water splitting
experiments. S. Hernández, C. Ottone and N. Russo performed and analyzed the chemical-
physical and optical characterizations of the bare and dye-sensitized BiVO4 samples. C. Ottone
experimentally determined the HOMO and LUMO energy levels of the dyes. K. Tolod did the
photoelectrochemical tests. S. Proto and S. Castillón designed and developed the organic
synthetic routes to obtain the core-substituted NDIs. S. Proto and C. Godard optimized the
grafting procedure of the dyes on BiVO4 powders. S. Proto and M. Díaz de los Bernardos
performed FT-IR, NMR, MS and TG characterizations under the supervision of C. Godard. A.
Solé-Daura and J. J. Carbó performed and analyzed the DFT calculations. G. Saracco and C.
Claver ideated and planned the research work. All authors discussed and analyzed the results,
contributed to writing and approved the final manuscript.
References
1.
2.
J. Barber, Chemical Society Reviews, 2009, 38, 185-196.
N. S. Lewis and D. G. Nocera, Proceedings of the National Academy of Sciences, 2006,
103, 15729-15735.
3.
4.
C. Acar, I. Dincer and G. F. Naterer, International Journal of Energy Research, 2016, 40,
1449-1473.
R. Abe, Journal of Photochemistry and Photobiology C: Photochemistry Reviews, 2010,
11, 179-209.
5.
6.
Y. Park, K. J. McDonald and K.-S. Choi, Chemical Society Reviews, 2013, 42, 2321-2337.
S. Hernández, G. Gerardi, K. Bejtka, A. Fina and N. Russo, Applied Catalysis B:
Environmental, 2016, 190, 66-74.
7.
L. Xia, J. Bai, J. Li, Q. Zeng, L. Li and B. Zhou, Applied Catalysis B: Environmental, 2017,
204, 127-133.
8.
9.
X. Meng and Z. Zhang, Journal of Molecular Catalysis A: Chemical, 2016, 423, 533-549.
A. Kudo, K. Omori and H. Kato, Journal of the American Chemical Society, 1999, 121,
11459-11467.
10.
D. K. Zhong, S. Choi and D. R. Gamelin, Journal of the American Chemical Society, 2011,
133, 18370-18377.
11.
12.
B. Zhou, J. Qu, X. Zhao and H. Liu, Journal of Environmental Sciences, 2011, 23, 151-159.
D. Ressnig, R. Kontic and G. R. Patzke, Materials Chemistry and Physics, 2012, 135, 457-
466.
13.
S. Obregón, A. Caballero and G. Colón, Applied Catalysis B: Environmental, 2012, 117–
118, 59-66.
14.
15.
G. Xi and J. Ye, Chemical Communications, 2010, 46, 1893-1895.
C. Martinez Suarez, S. Hernández and N. Russo, Applied Catalysis A: General, 2015, 504,
158-170.
16.
L. Wenjun, W. Zhiqiang, W. Lijuan, L. Zhaosheng, Y. Tao and Z. Zhigang, Journal of Physics
D: Applied Physics, 2010, 43, 405402.
17.
18.
19.
G. Li, D. Zhang and J. C. Yu, ChemInform, 2008, 39, no-no.
K. Tolod, S. Hernández and N. Russo, Catalysts, 2017, 7, 13.
B. Anke, M. Rohloff, M. G. Willinger, W. Hetaba, A. Fischer and M. Lerch, Solid State
Sciences, 2017, 63, 1-8.
19

Green Chemistry
Page 20 of 32
View Article Online
DOI: 10.1039/C7GC00125H
20.
21.
22.
H. Ye, J. Lee, J. S. Jang and A. J. Bard, The Journal of Physical Chemistry C, 2010, 114,
13322-13328.
J. Quiñonero, T. Lana-Villarreal and R. Gómez, Applied Catalysis B: Environmental, 2016,
194, 141-149.
S. M. Thalluri, S. Hernández, S. Bensaid, G. Saracco and N. Russo, Applied Catalysis B:
Environmental, 2016, 180, 630-636.
23.
24.
D. Pei and J. Luan, International Journal of Photoenergy, 2011, 2012.
W. J. Youngblood, S.-H. A. Lee, K. Maeda and T. E. Mallouk, Accounts of chemical
research, 2009, 42, 1966-1973.
25.
Y. Zhao, J. R. Swierk, J. D. Megiatto, B. Sherman, W. J. Youngblood, D. Qin, D. M. Lentz,
A. L. Moore, T. A. Moore, D. Gust and T. E. Mallouk, Proceedings of the National
Academy of Sciences, 2012, 109, 15612-15616.
26.
J. R. Swierk, D. D. Méndez-Hernández, N. S. McCool, P. Liddell, Y. Terazono, I. Pahk, J. J.
Tomlin, N. V. Oster, T. A. Moore, A. L. Moore, D. Gust and T. E. Mallouk, Proceedings of
the National Academy of Sciences, 2015, 112, 1681-1686.
27.
28.
29.
30.
31.
H. Yu, R. Shi, Y. Zhao, G. I. N. Waterhouse, L. Z. Wu, C. H. Tung and T. Zhang, Advanced
Materials, 2016, 28, 9454-9477.
D. Tang, H. Zhang, H. Huang, R. Liu, Y. Han, Y. Liu, C. Tong and Z. Kang, Dalton
Transactions, 2013, 42, 6285-6289.
D. Hidalgo, S. Bocchini, M. Fontana, G. Saracco and S. Hernandez, RSC Advances, 2015,
5, 49429-49438.
S. M. Thalluri, C. Martinez Suarez, S. Hernández, S. Bensaid, G. Saracco and N. Russo,
Chemical Engineering Journal, 2014, 245, 124-132.
D. D. Perrin, W. L. F. Armarego and D. R. Perrin, Purification of laboratory chemicals,
Pergamon Press, Oxford; New York, 1980.
32.
33.
K. Moller and T. Bein, 1998, DOI: 10.1021/cm980243e.
J. Rochford, D. Chu, A. Hagfeldt and E. Galoppini, Journal of the American Chemical
Society, 2007, 129, 4655-4665.
34.
35.
36.
J. Zhang, C. Yu, L. Wang, Y. Li, Y. Ren and K. Shum, Scientific Reports, 2014, 4, 6954.
V. V. Pavlishchuk and A. W. Addison, Inorganica Chimica Acta, 2000, 298, 97-102.
G. W. T. M. J. Frisch, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, G.
Scalmani, V. Barone, B. Mennucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li, H. P.
Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Sonnenberg, M. Hada, M. Ehara, K.
Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, T.
Vreven, J. A. Montgomery, Jr., J. E. Peralta, F. Ogliaro, M. Bearpark, J. J. Heyd, E.
Brothers, K. N. Kudin, V. N. Staroverov, R. Kobayashi, J. Normand, K. Raghavachari, A.
Rendell, J. C. Burant, S. S. Iyengar, J. Tomasi, M. Cossi, N. Rega, J. M. Millam, M. Klene,
J. E. Knox, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann, O.
Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, R. L. Martin, K. Morokuma, V.
G. Zakrzewski, G. A. Voth, P. Salvador, J. J. Dannenberg, S. Dapprich, A. D. Daniels, O.
Farkas, J. B. Foresman, J. V. Ortiz, J. Cioslowski and A. D. J. Fox, Gaussian 09, revision
C.01, 2010, Gaussian, Inc.: Wallingford CT.
37.
38.
39.
40.
41.
J. P. Perdew, Physical Review B, 1986, 33, 8822-8824.
A. D. Becke, Physical Review A, 1988, 38, 3098-3100.
J. P. Perdew, K. Burke and M. Ernzerhof, Physical Review Letters, 1996, 77, 3865-3868.
J. P. Perdew, K. Burke and M. Ernzerhof, Physical Review Letters, 1997, 78, 1396-1396.
M. M. Francl, W. J. Pietro, W. J. Hehre, J. S. Binkley, M. S. Gordon, D. J. DeFrees and J. A.
Pople, The Journal of Chemical Physics, 1982, 77, 3654-3665.
42.
W. J. Hehre, R. Ditchfield and J. A. Pople, The Journal of Chemical Physics, 1972, 56, 2257-
2261.
43.
44.
P. C. Hariharan and J. A. Pople, Theoretica chimica acta, 1973, 28, 213-222.
P. Pulay and G. Fogarasi, The Journal of Chemical Physics, 1992, 96, 2856-2860.
20

Page 21 of 32
Green Chemistry
View Article Online
DOI: 10.1039/C7GC00125H
45.
46.
E. Cancès, B. Mennucci and J. Tomasi, The Journal of Chemical Physics, 1997, 107, 3032-
3041.
C. Thalacker, C. Röger and F. Würthner, The Journal of Organic Chemistry, 2006, 71,
8098-8105.
47.
48.
49.
S. Ding and N. Jiao, Angewandte Chemie International Edition, 2012, 51, 9226-9237.
B. Basu and S. Kundu, New Journal of Chemistry, 2014, 38, 3367-3370.
T. Saison, N. Chemin, C. Chanéac, O. Durupthy, V. Ruaux, L. Mariey, F. Maugé, P.
Beaunier and J.-P. Jolivet, The Journal of Physical Chemistry C, 2011, 115, 5657-5666.
S. Bensaid, C. Ottone Melis, S. Hernández, M. Armandi, S. Esposito, G. Saracco and B.
Bonelli, Chemical Engineering Journal, 2017, 311, 143-152.
50.
51.
52.
G. Zhang and C. B. Musgrave, The Journal of Physical Chemistry A, 2007, 111, 1554-1561.
A. Iwase, Y. H. Ng, Y. Ishiguro, A. Kudo and R. Amal, Journal of the American Chemical
Society, 2011, 133, 11054-11057.
53.
54.
55.
D. Hidalgo, R. Messina, A. Sacco, D. Manfredi, S. Vankova, E. Garrone, G. Saracco and S.
Hernández, International Journal of Hydrogen Energy, 2014, 39, 21512–21522.
S. Hernández, M. Tortello, A. Sacco, M. Quaglio, T. Meyer, S. Bianco, G. Saracco, C. F.
Pirri and E. Tresso, Electrochimica Acta, 2014, 131, 184-194.
Y. H. Ng, A. Iwase, A. Kudo and R. Amal, The Journal of Physical Chemistry Letters, 2010,
1, 2607-2612.
56.
57.
S. V. Bhosale, C. H. Jani and S. J. Langford, Chemical Society Reviews, 2008, 37, 331-342.
H. E. Katz, A. J. Lovinger, J. Johnson, C. Kloc, T. Siegrist, W. Li, Y. Y. Lin and A. Dodabalapur,
Nature, 2000, 404, 478-481.
58.
59.
R. Abe, K. Shinmei, N. Koumura, K. Hara and B. Ohtani, Journal of the American Chemical
Society, 2013, 135, 16872-16884.
M. Higashi, R. Abe, T. Takata and K. Domen, Chemistry of Materials, 2009, 21, 1543-
1549.
21

Green Chemistry
Page 22 of 32
View Article Online
DOI: 10.1039/C7GC00125H
Figure 1. Synthetic strategy for the preparation of substituted-Naphtalenediimides. Reagent conditions: (i)
4-Aminopyridine/DMF, 135 ºC / 2 4h; (ii) Tribromocyanuric acid/oleum, 140 ºC / 24 h: (iii) 2,6-
Diisopropylanyline/CH3COOH, 120 ºC / 25 min; (iv) 10 equiv. of 4-Aminobenzoic acid/DMF, 135ºC/2h; (v)
40 equiv. of 4-Aminobenzoic acid /DMF, 135 ºC / 5 h; (vi) 40 equiv. of 4-Aminobenzoic acid/DMF, 150 ºC /
2 h.
153x76mm (150 x 150 DPI)

Page 23 of 32
Green Chemistry
View Article Online
DOI: 10.1039/C7GC00125H
Figure 2. Surface functionalization technique
93x51mm (220 x 220 DPI)

Green Chemistry
Page 24 of 32
View Article Online
DOI: 10.1039/C7GC00125H
Figure 3. FE-SEM images of (a) pristine BiVO4 powder and impregnated with (b) dye 1 and (c) dye 4.
79x189mm (300 x 300 DPI)

Page 25 of 32
Green Chemistry
View Article Online
DOI: 10.1039/C7GC00125H
Figure 4: Physicochemical characterization: (a) Xꢀray diffraction patterns and (b) UV−Vis spectra in total
reflectance of the pristine BiVO4 powder and impregnated with dyes 1 to 4.
154x64mm (300 x 300 DPI)

Green Chemistry
Page 26 of 32
View Article Online
DOI: 10.1039/C7GC00125H
Figure 5. High resolution (a) C 1s, (b) O 1s, (c) N 1s, (d) Bi 4f and (f) V 2p XPS spectra of the pristine BiVO4
powder and after the impregnation with dyes 1 to 4.
209x251mm (300 x 300 DPI)

Page 27 of 32
Green Chemistry
View Article Online
DOI: 10.1039/C7GC00125H
Figure 6. Photocatalytic water oxidation measurements. (a) Picture of the Clark-cell system in the lab and
comparison of the (b) oxygen evolution curves of the pristine BiVO4 and impregnated with the different dyes
and (c) the Clark-type cell parameters of oxygen evolution rate (RO2) and maximum dissolved oxygen
(DO*).
188x159mm (300 x 300 DPI)

Green Chemistry
Page 28 of 32
View Article Online
DOI: 10.1039/C7GC00125H
Figure 7. Energy levels of the different dyes and BiVO4, including BP86 isodensity surfaces for the computed
HOMO and LUMO of the different dyes. Hydrogen atoms are omitted for clarity.
296x181mm (300 x 300 DPI)

Page 29 of 32
Green Chemistry
View Article Online
DOI: 10.1039/C7GC00125H
Figure 8. Conceptual representation of the mechanisms of UV-Vis light absorption, charge carriers
generation, transfer and separation in the BiVO4 + dye photocatalytic systems. (a) Z scheme-type
mechanism for the photochemical water oxidation reaction under the sacrificial electrons acceptor (AgNO3)
of the photocatalysts in the powder form. (b) Combined mechanism of charges generation, separation and
reaction in the overall water splitting photoelectrochemical cell exploiting the BiVO4 + dye 3 photoanode.
160x52mm (300 x 300 DPI)