D.K. Bora, A. Braun, E.C. Constable, “In rust we trust”.

Hematite the prospective inorganic backbone for artificial

photosynthesis, Energy Environ. Sci., 2013,6, 407-425.
http://pubs.rsc.org/en/content/articlelanding/2012/EE/C2EE23668K

“In rust we trust”. Hematite - the prospective inorganic backbone for

artificial photosynthesis
Debajeet K. Bora,*a,b,c Artur Braun a and Edwin C. Constable b
5

The search for affordable high performance electrode materials in photoelectrochemical hydrogen production by solar water
splitting is an on-going quest. Hematite is a photoanode material with an electronic band gap suitable for efficient absorption of
visible light in a photoelectrochemical cell (PEC). Although its poor electronic structure makes hematite a controversial candidate
for PEC, it remains promising because it is an earth abundant, chemically stable and low cost material – necessary prerequisites for
10 PEC to become a competitive cost-efficient solar fuel economy. In addition to reviewing some recent PEC research on hematite and
its relevant physical and chemical characteristics, we show how hematite obtained by a low cost synthesis can be refined by
hydrothermal treatment and further functionalized by coating with phycocyanin, a light harvesting protein known for photosynthesis
in blue-green algae.

The components of photo-electrochemical cells (PEC) for solar hydrogen production by water splitting must be based rather on “no-cost”
15 materials than on low-cost materials. Hematite is in many aspects an attractive photoanode alternative, but extensive materials
engineering is required before hematite, “rust” can compete with the performance of conventional systems. But there is encouraging
progress going on in making hematite a high performance PEC photoanode. Meanwhile, functionalization of hematite with phycocyanin
proteins from another potential low cost source, blue green algae, shows that nano-bio hybride architectures are possible where hematite
provides both a functional and a structural backbone for artificial photosynthesis. C – Phycocyanin, a light harvesting antenna molecule
20 found in cyanobacteria has been taken part in water splitting reaction since 3 billion years ago. The purpose of this molecule is to funnel
the light harvested energy into the reaction centre. Inspired by this function, we have utilized this molecule to build a hybrid integrated
photoanode system from nanostructured hematite and c-phycocyanin. The functionality and stability of the resulting photoanode system
makes it possible to employ it for photoelectrochemical solar splitting. Since this approach is based on simple dip coating fabrication
steps, scale-up could make it promising for large scale applications.
25

Introduction
The increased demand for clean energy sources is driving
30 research for the development of alternative energy technologies.
Based on the requirement for clean, preferably carbon-neutral
technologies, hydrogen is considered as an ecologically benign
green fuel with high energy density.2 Plants and cyanobacteria
produce hydrogen through the water splitting machinery of the
35 photosynthetic process.1 This photon-driven water-splitting
method provides inspiration for current approaches to biomimetic
or low-tech hydrogen generating technologies. The basic energy
conversion scheme in the photon-driven water-splitting method is
the transformation of light energy into chemical energy. Artificial
40 inorganic photonic systems are expected to play an important role Figure 1. Basic building principle of an electrochemical cell with wired
in the sustainable production of hydrogen by photoanode and photocathode in aqueous electrolyte and illuminated by
photoelectrochemical water splitting using sunlight as a primary 50 sunlight.
energy source. Photoelectrochemical cells3 mimic the processes
of photosystem I and II4 and work on simple principles of Here we focus on the use of hematite as a photoanode in PECs
45 semiconductor electrochemistry.5 A simplistic architecture of a for solar hydrogen generation by water splitting. Note, that, the
typical photoelectrochemical cell is presented in Figure 1. photoanode is the oxygen evolving electrode, whereas the
hydrogen is generated at the photocathode.
55 Photoelectrochemistry at the interface between inorganic and
organic components is chosen as an alternative to increase the
efficiency of a photoelectrochemical cell. The system acts like an
“artificial photosynthetic device” whereby it mimics the functions
 of photosystem II and Photosystem I. With the current advent of relatively benign.14 An additional factor in favor of iron-based
nanotechnology and surface functionalization strategies it is and hematite materials is the low per unit cost.
possible to develop such hybrid photoelectrode systems based on 60 Hematite photoanodes were first investigated in PECs by Bard
nanobiointerface concepts. Based on this, recently many systems and Hardee,15 who found hematite to be very stable in neutral and
5 have been developed. In our case, we have developed a novel alkaline electrolytes under PEC operating conditions. The low
kind of photoelectrode system at the interface of nanostructured efficiencies observed were interpreted using Gärtner theory for
hematite electrodes and the light harvesting membrane protein C- semiconductor photoresponse,16 and it was proposed that poor
Phycocyanin. In this perspective article, we discuss some 65 light absorption capability and a short hole diffusion length
effective surface functionalization strategies for increasing the limited the efficiency. Hematite modified by surface pretreatment
10 efficiency of hematite-based photoelectrode systems. Our has been shown to have improved properties and it was observed
functionalized hematite system showed two fold increases in that photocurrent onset was delayed due to a very low Faradic
photocurrent in comparison to pristine ones by both hydrothermal rate constant for water oxidation chemistry17. Inspired by these
and bioconjugation modification techniques. These strategies can 70 observations, we developed hematite-based hybrid nanostructures
be further extended to other metal oxide system such as WO3, or for enhancing photocurrent through the surface functionalization
15 ZnO. and hydrothermal modification of pristine hematite electrodes.

Electronic structure of hematite and iron oxides

Photoelectrochemical cells and the choice of
Iron oxide is a seemingly simple compound, but there exists a
hematite
75 manifold of iron oxide phases, which are strongly correlated
The photoelectric effect was discovered by Becquerel in 1839.6 In electron systems with complex electronic structures. Furthermore,
20 semiconductor electrodes, this effect is more prominent because the surfaces of these oxides may interact with the ambient
the photogenerated electron-hole pair drives the redox reaction. environment/air and further complicate the structure, for example
The presence of the electric field facilitates the photocurrent or by hydroxylation. Under PEC conditions in the favored alkaline
photovoltage, once charge carriers of opposite signs are 80 electrolyte, the electrode is exposed to water and additional
generated.7 The semiconductor-electrolyte interface is critical and hydroxyl groups, with which the iron oxide surfaces form even
25 the potential differences between the two phases (semiconductor more complex phases. The complexity of the electronic structure
and liquid) are controlled by the space-charge layer, considered to is reflected by the catalytic properties and transport properties.11c
be part of electrical double layer.7 The photocurrent is determined Iron oxides are trivially summarized as rusts, particularly when
by the electric field in this space-charge layer. 85 the iron oxide is formed in the presence of water or humidity.
The PEC is an application of this concept and was originally Certainly, we must take this case into account in PEC conditions.
30 developed by Fujishima and Honda in 1972,8 who employed an It belongs to the one of the different phases of iron oxide or rust.
n-type TiO2 electrode (photoanode) for the photoelectrolysis of Rust itself is nothing but hydrated ferric oxide or iron(III)
water at an efficiency of 0.1%. Only a short time later, Calvin oxyhydroxides. We cannot utilize rust directly for solar water
designed a PEC cell in which the electrode surface was optically 90 splitting because of slow electronic conduction with polaron
sensitized with dye molecules.9 In this sensitized photoelectrode, hopping. 11d
35 the generation of the photocurrent is also determined by the Hematite possesses a corundum structure 18a with Fe(III) cations
space-charge layer at the semiconductor–dye–electrolyte in a slightly distorted octahedral environment. Hematite has
interface. The photoresponse in a PEC cell is generally optical absorption bands in the UV and near IR region due to
determined by the potential barrier generated at the 95 crystal field and CT transitions.19a The band gap ranges from 1.9 -
semiconductor–electrolyte interface through thermodynamic 2.2 eV with an indirect phonon-assisted band-gap transition.20
40 equilibration. This equilibrium is established by the equalization The LMCT absorption at 400 nm combined with pair excitation
of Fermi levels, where the majority charge carrier (an electron in processes at 485 – 500 nm results in the evolution of a broad
case of n-type semiconductors) is transferred from the electrode absorption band 400 – 600 nm.21 The light absorption of
to the electrolyte.10 100 hematite in the long wavelength region is far larger than that of
Gerischer has proposed a model for the kinetics and energetics the closely related phases such as maghemite and magnetite.18b
45 of electron-transfer across semiconductor-electrolyte junctions.7 In hematite nanoparticles, it has been observed that the relative
In addition to understanding the physical chemistry of the PEC intensities of different spectral regions change with particle
cell, significant efforts have been devoted to the development and size.19b, 19c This is believed to be due to enhancement of the pair
optimization of sustainable photoelectrode materials. Numerous 105 excitation in larger particles together with a redshift of the CT
semiconducting metal oxide photoelectrodes have been tested for band.22
50 photoelectrochemical applications,11a and of these hematite (α- The electronic structure of pristine hematite comprises an
Fe2O3) is attractive on the basis of its stability in electrolyte and empty conduction band with Fe 3d character and a valence band
its absorption spectrum.11b. originating from Fe 3d(2t2g) and O 2p orbitals.18b When hematite
One factor favoring hematite semiconductor electrodes is the 110 is used as a photoanode in a PEC, the interaction with aqueous
high abundance of iron with an average estimate of 2 to 3% in electrolyte can influence its electronic structure, especially at the
55 sedimentary rocks rising to 8.5% in basalt and gabbro, making it semiconductor liquid junction. Even in the absence of an external
the fourth most abundant element in the earth’s crust.12 electric bias, a chemically induced surface potential gradient can
Furthermore, iron compounds are generally biocompatible13 and be generated.23a For a long time, electrochemical impedance
 spectroscopy has been a helpful analytical tool in order to probe t2g-eg doublet from hematite, and the new state from
charge carrier dynamics and electronic structure of photo- anodization, which most likely corresponds to a hydroxyl or
electrochemical systems. One recent work on hematite is found in oxo-hydroxyl group at the surface of the hematite. Hence, at the
ref. 23b. 23b hematite surface we have indeed a rust-like species. In a recent
5 In studying the electronic structure of hematite, oxygen K-edge 45 study, the electronic structure of a working Si-doped hematite
soft X-ray spectroscopy has proven to be useful. The photoanode system was studied in-situ with O K-edge NEXAFS
spectroscopic sensitivity of the oxygen K-edge depends on the spectroscopy. During this study, we have observed two different
local bonding. The pre-peaks of the K-shell absorption edge are types of O K-edge NEXAFS spectra in the dark and light
governed by the 3d components in the unoccupied p-d hybrid condition at different applied bias which suggested the presence
10 wave functions.24 Photoemission spectroscopy confirms that 50 of two types of holes which were assigned to transitions into the
hematite is a charge transfer insulator, i.e. hematite has a band CTB and UHB, as shown in the lower panel of Figure 2.31b Being
gap between the O 2p and Fe 3d bands.25 The oxygen NEXAFS a small polaron conductor, hematite becomes conductive at high
spectrum of hematite originates from transitions from the O 1s temperatures32a, 32b and various models have been proposed to
core orbitals to Fe (3d-4s) and Fe 4p-like bands or molecular describe these properties. Bosnian and Van Daal developed a
15 orbitals having some O 2p character.26 The doublet in the pre- 55 hopping conduction model in a narrow band gap semiconductor
edge region at 528 - 530 eV, originates from transitions of Fe to elucidate the charge transport mechanism of various transition
3d–O 2p states with t2g and eg orbital symmetry.27 NEXAFS metal oxides,33 and also studied the frequency dependence of
spectroscopy at the O K-edge was also used to investigate the conductivity at low temperatures. In contrast, Gardner used DC
band edge electronic structure of bulk and nanoscale hematite.28 conductivity measurements to show that below 1073 K, hematite
20 The band gap widening in iron oxide nanoparticles is potentially 60 behaves as an extrinsic semiconductor with the conductivity very
important for application in solar cells.29,30 Recently, ex-situ X- sensitive to stoichiometry.34 Hematite is paramagnetic above 956 K,
ray absorption studies of hematite nanoparticulate films with a weak ferromagnetic character at room temperature. The Morin
anodized at 600 mV anodic bias produced an extra transition in transition temperature to an antiferromagnetic state occurs at 260 K
the upper Hubbard band (UHB), which we interpreted as being due to magnetic compensation of sublattices.35 Hematite
25 related to a highly oxidized Fe species.31a Figure 2 shows the 65 nanoparticles with particle sizes of around 8 nm show
NEXAFS spectra of pristine 4 layer film (600 nm thin) and a 10 superparamagnetic relaxation at room temperature.36
layer films (1500 nm) anodized at 600 mV.
Hematite in the PEC environment
The photoanodes in a photoelectrochemical cell (PEC) comprise
thin films of hematite on a conducting and typically transparent
70 substrate, such as FTO on glass. Hematite films and nanoparticles
have controversial semiconducting behaviour because of the
dependence on thickness and surface properties. The surfaces are
highly influenced by pre-treatment processes which may partially
convert hematite into magnetite.17, 31a Nevertheless, the stability of
75 hematite in solution at pH > 4 combined with an absorption having
energy greater than the band gap makes it suitable for
photoelectrochemical water oxidation.15,37,38 Hematite has the
disadvantages of low quantum efficiency and high resistivity.
Processes such as charge mobility and charge transfer across the
80 interface39 play a critical role in these intrinsic limitations.
Topography plays a role, and hematite thin films with needle-like
nanoparticles on conducting glass substrates were found to be more
efficient than the films of sintered particles.40
Photogenerated holes form narrow bands with 3 d character and
85 this is a major factor affecting the behaviour of hematite when in
contact with aqueous electrolytes.39 Decreased efficiency of hematite
Figure 2. Top – Oxygen K-edge NEXAFS spectra obtained at BESSY-II in the long wavelength region is due to limitations of carrier diffusion
30 ex-situ from 600 nm pristine hematite (dotted line) and 1500 nm hematite
anodized at 600 mV (solid line), and their difference spectrum (middle length. The electron-hole pairs are located deep in the semiconductor
panel). The third peak at 532.85 eV is the result of anodization in KOH bulk and the low mobility combined with the small diffusion length
and not present in pristine hematite. Bottom – NEXAFS spectrum 90 of holes results in a high probability of recombination. On the other
recorded in-situ in a liquid cell at the Advanced Light Source under 500 hand, efficiency is greater at short wavelengths due to the generation
35 mV anodic bias in KOH electrolyte and 1.5 AM illumination.
of the electron-hole pair in the near-surface region where a space
The extra peak at 533 eV is formed upon anodization at 600 mV. charge layer exists. In this case the photogenerated holes do not have
An anodization potential of 200 mV was not sufficient to create to diffuse a long distance to oxidize the water molecule. In the water
this peak. The extra transition reveals that surface electronic 95 oxidation process, the charge transfer kinetics are slow due to energy
states can be altered by the PEC operating conditions. The mismatching between the acceptor d orbitals of hematite and donor p
40 difference spectrum at the bottom in Figure 2 shows clearly the orbitals of the O2-OH– redox couple.39 The photocurrent can
 drastically decrease when electronic defects are located at or near the diffusion length of lithium ions. The chemical oxidation of
surface where the holes are supposed to enter the electrolyte.41 Some 40 organics on hematite surfaces is actively investigated, and
studies have suggested that poor photocurrents also arise as a result of hematite has been studied as a heterogeneous catalyst for the
grain boundaries, surface states or crystalline imperfection acting as a degradation51, 52 and alkylation 53 of phenols. In the
5 recombination sites for carrier trapping.42-45 photodegradation of azo dyes such as orange II on hematite
surfaces by photo oxidative processes, it was found that the
Hematite nanoparticles 45 catalytic oxidation was sustainable in a cyclic repetitive way,
with the rate strongly depending on the sintering temperature of
Hematite nanoparticles are finding increasing use in a wide the nanoparticles.54,55 The photoelectrochemical properties of
range of applications due to their physicochemical properties. hematite films also depend on the sintering temperature.56 High
Photoanode films have been fabricated for photoelectrochemical temperature treatment of nanoparticles can be used to reduce the
10 water oxidation using 5 - 16 nm nanoparticles.46 Nanoparticles
50 insulating maghemite phase which drastically decreases the
were prepared by thermal decomposition of iron oleate performance of hematite films.57 Structural changes of hematite
precursors. Only the smallest particles of 5 nm constitute pure nanoparticles induced by high temperature treatment has been
hematite, with other sizes being mixed phases, although the phase studied with Fe K-edge XANES spectroscopy. The intensities of
content could be varied with processing temperature. the two peaks at 7129 eV and 7132.7 eV (indicated by the dashed
15 Nanoparticles of 12 nm average diameter showed a maximum
55 lines in the left panel in Figure 3) is systematically redistributing,
photocurrent density as a result of reduced electron-hole and the pre-edge peaks (deconvoluted for the sample heated at
recombination due to optimal matching of nanoparticle size with 550°C in the lower right panel in Figure 3) intensities are
the hole diffusion length. Hematite nanoparticles have been changing as well depending on heat treatment temperature. The
electrodeposited on the surface of TiO2 nanotube arrays obtained t2g peak at around 7113 eV decreases with annealing temperature
via anodization of Ti foil,47 and tuning of PEC activity was due to changes in oxidation state and site symmetry (Figure 3).58
20
60
achieved by varying the coverage of hematite nanocrystals Figure 4 displays the evolution of nanoparticle size with heat
through pretreatment with aqueous iron (III) salts. treatment temperature from 250 to 900°C validated by XRD and
light scattering techniques.

Figure 3. left panel - Fe K-shell absorption spectra of hematite

25 nanoparticles heated from 250°C to 900°C. lower right panel -
Deconvoluted pre-edge peak of the XANES spectrum for 550°C annealed
iron oxide nanoparticles (size ~ 100 nm). Upper right panel - Relative
spectral weight of the t2g peak versus annealing temperature. 65 Figure 4. Evolution of particle size (top, from light scattering) and
crystallite size (bottom, from XRD using Scherrer’s equation) of iron
Photocatalytic water oxidation by hematite nanocrystals has oxide nanoparticles annealed from 250°C to 900°C.
30 been studied and a high photocatalytic O2 evolution rate of 1072
mmol h-1 g-1 was obtained in comparison to bulk and sonicated Infrared spectroscopy studies centering on the vibrational
hematite.48 In addition to PEC applications, the high specific mode at 690 cm-1, confirmed the presence of maghemite phases
charge storage capacity of hematite nanoparticles has led to their 70 in hematite nanoparticles.59 The overall structural features of
being incorporated into lithium ion batteries. It was found that 30 hematite nanoparticles are controlled by both the particle size and
35 nm nanocrystals had an optimal discharge capacity. Mesoporous growth kinetics,60-63 and significant differences exist between
hematite anodes with a specific capacity of 1360 mA h g-1 have bulk material and the nanoscale. For instance, in heat capacity
been used in lithium ion batteries.49,50 Also here, morphology measurements of 13 nm hematite nanoparticles (T =1.5 to 350 K)
plays an important role in nanocrystals due to the reduction in the 75 using semi-adiabatic and adiabatic calorimetry, higher heat
 capacities obtained than the bulk, and were attributed to a greater precursors by methods including sonochemistry89,
ratio of surface water to bulk hematite.64 mechanochemistry,90 hydrolysis, 91 thermolysis 92 and co-
We have learnt that the structure, electronic and optical precipitation.93 Co-precipitation in the bulk is an economic and
properties and transport properties of hematite depend on the size. 60 versatile technique that may be used to synthesize large amounts
5 This naturally includes the magnetic properties. Bulk hematite is of materials with different compositions and particle sizes.
weakly ferromagnetic at room temperature, but below the Morin Nanoparticles obtained in batch co-precipitation have a broad size
transition temperature at 260 K, a first order spin reorientation distribution and a greater control of the synthesis conditions is
becomes important, 65-68 and there is a loss in net magnetic required. 94 Near monodisperse nanoparticles have been produced
moment and an antiferromagnetic transformation. On the other 65 by thermal decomposition of an iron-cupferron complex in
10 hand, in nanocrystalline hematite (size ~ 200 nm) no spin octylamine.95 Organometallic complexes of iron have also been
reorientation is observed upon cooling, and ferromagnetic used as precursors and an Fe(CO)5, oleic acid and trimethyl oxide
behavior persists to T = 2K.69-74 mixture in octyl ether has been used for maghemite synthesis.96
The use of nanoscale hematite for solar energy applications is Magnetite can be prepared from [Fe(acac)3] in oleic acid, 1,2-
largely determined by its electronic properties, particularly by the 70 hexadecanediol, olyelamine and diphenyl ether.97 The reaction of
15 widening of the band gap energy and to a lesser extent by ferrocene with hydrazine in PEG-2000 in a surfactant assisted
quantum confinement effects.75 Due to strongly localized Fe 3d solvothermal synthesis leads to ferromagnetic Fe3O4 nanodiscs
states, it is very unlikely that quantum confinement effects will be which can be used in high density recording media.98
observed in nanoscale hematite. 76-77 It has been concluded that While hematite nanoparticles can be prepared from molecular
small particle size may influence the energy positions of the 75 precursors by a range of techniques,99-103 simple and affordable
20 electronic states which define the semiconductor band gap 78, yet iron(III) salts, such as chloride or nitrate can also be used for the
a recent study shows no change in the electronic properties of synthesis of iron oxides and ferrites, but typically do not lead to
hematite 79. Here, oxygen K-edge absorption and emission nanomaterials in the absence of additives. Combustion flame
spectroscopies have been employed to study valence-band and synthesis has been used to produce hematite nanoparticles 103a.
conduction-band electronic structures of nanoscale hematite with 80 Surfactants can act as surface-capping groups and stabilize iron
25 different size ranges. Similar techniques have been used to study oxide nanomaterials. The morphology of iron oxide
the electronic properties of bulk hematite.80-83 A strong nanomaterials may be controlled by the choice of precursors and
correlation between the hematite optical band gap and the energy additives. A versatile, low cost and scalable dip coating technique
separation between the onset of emission and absorption spectra has been developed recently for the synthesis of iron oxide
was obtained and it was found that nanoparticles down to 8 nm 85 nanoparticles through a benign chemistry route.104 Unlike some
30 did not show any changes in the band gap controlled by variation processes where large aggregates or products with variable
in size.84 oxidation states or matrix inclusion are obtained, the new
technique is free from these problems. A main characteristic of
Synthesis of hematite nanoparticles this benign route is the lower temperature of the thermal
90 treatment. The technique also yields a stoichiometric hematite
Top down approach
product with small particle size and high homogeneity.
Mechanical milling is one top down method for the synthesis of Biomolecules are effective templates for the growth and self-
35 hematite nanoparticles. Hematite-maghemite composites may be assembly of nanostructures,105 and the hydrothermal reaction of
prepared by mechanical milling in air at 300 RPM in hardened iron salts with amino acids leads to the formation of hematite
steel vials using hardened steel balls.85 Planetary ball milling of 95 nanocubes.106 We have modified this method to obtain hematite
hematite powders in air or ethanol has also been investigated.86 nanoarchitectures with well-defined geometries and with surface
X-ray diffraction with Rietveld structure refinement, and modification that can be further utilized for functionalization of
40 Mössbauer spectroscopy showed partial transformation from α- the nanoparticle surface and fabrication of bioconjugates. The
Fe2O3 to a spinel phase after extensive milling in air, while only amino acid is crucial for the controlled crystallization of hematite
the α-Fe2O3 phase was obtained when the material was dispersed 100 nanomaterials.
in ethanol. Average grain sizes for the nanostructured hematite
milled in air were always smaller than those milled in ethanol, Bare hematite in photoelectrochemical cells
45 reaching a minimum of about 11 nm. Pure hematite particles of
Hematite nanocrystals have been used to prepare photoanodes for
average size of 14 nm in diameter have been obtained from ball
the photooxidation of water. Thin films of the nanoparticles (bare
milling in a vertical vibrating mixer mill,87 and Mössbauer and
or pristine hematite) are prepared by deposition methods. We will
magnetization measurements indicate the existence of strong
interparticle interactions that can be varied by particle dispersion. 105 now discuss the development and optimisation of photoanodes
The milled nanoparticles had surface areas of around 15 m2 g–1 using bare hematite (without any other surface modification).
50

and degrees of crystallinity from 9 to 50%. In another study, the Pristine Hematite Films for PEC applications
effect of rotational speed was investigated and a minimum
Thermal decomposition of solid FeCl3⋅6H2O on FTO is a simple
crystallite size of 17.1 nm in diameter with a degree of
and low cost method for preparing hematite thin films.107 The
crystallinity of 9.37% was obtained at 600 rpm for 10 hours.88
110 study showed that the film obtained resembles a cactus-like field
55 Bottom up approach with 20 nm thick porous crystal plates as was evident from SEM
imaging.
Iron oxide nanoparticles have been prepared from molecular
 The particle size in nanocomposite hematite-SnO2 thin films Dip coated pristine hematite for PEC application
fabricated from homogeneously mixed precursor solutions may We present now a method for the deposition of hematite films
60
be controlled by varying the precursor composition and post-
using precursor materials derived from non-aqueous processing
deposition annealing conditions.108 Composite films with
of iron(III) salts and oleic acid. Films are deposited by dip
5 optimised particle sizes exhibit quantum confinement effects. coating of FTO coated glass in the precursor solution followed by
Sivula developed a solution-based colloidal deposition of sintering at 500°C for 2 hr (Figure 5).31(a)
mesoporous hematite thin films giving a water splitting 65 The as-synthesized films were studied with XRD (Figure 6);
photocurrent density of 0.56 mA cm-2,56 and a saturation
the major diffraction peaks arise from the dense SnO2 coating of
photocurrent density of 1 mA cm-2 was obtained before dark the glass substrate (0.4µm cassiterite) and only one strong peak
10 current onset. On annealing the hematite film, the nanoparticle due to hematite is observed (the (110) reflection in hexagonal
size was increased, drastically changing the optical properties of coordinates), in addition to weaker Bragg reflections
the film. From XPS and SQUID measurements, it was concluded corresponding to the (012), (104), (113), (024) and (300) planes.
70
that tin was doped into the hematite lattice by temperature
activated diffusion. Hematite thin films have also been prepared
15 by pulsed laser deposition from hematite targets.109 Films with a
thickness of 70 nm using SiO2 sphere template gave highest
photocurrent densities of 275µA cm-2 under AM 1.5 solar
irradiation. Spray pyrolysis of 0.11M FeCl3 in ethanol has also
been used for the preparation of hematite thin films giving a
20 maximum photocurrent density of 3.7 mA cm-2 under optimized
conditions.110 A photocurrent efficiency of 1.84% was reported at
an applied potential of 0.2V vs. the saturated calomel electrode
(SCE) at a pH of 14. In a different study, reactive ballistic
deposited hematite films, prepared by evaporative deposition of
25 Fe metal in oxygen onto a substrate showed an IPCE of 10% at
420 nm in 1M KOH.111
Hematite films can also be prepared by electrodeposition
which is a low cost, versatile and easily up-scalable technique.112
Films were deposited on glass substrates using a non-aqueous
30 precursor solution with subsequent thermal oxidation in air. High Figure 5: Synthesis procedure for hematite film by dip coating.30
current densities allow for deposition of more structured and
porous films and also yield a smaller size of the crystallite along A representative image of the morphology of the hematite
the crystallographically preferred 110 direction. 75 films as obtained by FESEM is shown in Figure 6 (inset in
Spin-coated hematite films on FTO obtained from polymeric bottom image) and consists of larvae shape nanostructures. The
35 precursors prepared by the condensation of citrate salts have also average diameter of the primary particles was 80 nm. Each
been investigated for photoelectrochemical water splitting.113 The nanostructure was the result of the incorporation and
precursors significantly influence the photoelectrochemical agglomeration of individual particles. Individual particles are
functionality of hematite films. For example, hematite films 80 fused to each other forming an interconnected 3D network with
obtained by the aerosol-assisted chemical vapour deposition of a electronic conducting paths facilitating charge transport.
40 [Fe6(PhCOO)10(acac)2(O)2(OH)2].3C7H8 exhibit a photocurrent The i-V curve of such a modified electrode in the dark and
density of 600 µA cm-2.40 Hematite films of differing thickness under irradiation is presented in Figure 6 (bottom). A
with a polycrystalline hexagonal structure were prepared by radio photocurrent density of 250 µA cm–2 is achieved at 1.23 V vs.
frequency magnetron sputtering,114 and showed good 85 RHE for FTO functionalised with one dip-coated hematite layer,
photocatalytic behaviour under visible light irradiation. The while no dark current is observed up to about 1.6 V where
45 thicker films showed enhanced photocatalytic action due to the oxygen evolution begins.
higher surface roughness which provided more reactive sites for The dip coated films have a very rough surface. Films obtained
photocatalysis. In addition to the photocatalytic activity, the film by spin coating from the same precursor and same heat treatment
thickness also influences the optical and photoelectrochemical 90 procedure are much smoother and glossy, but exhibit virtually no
properties of the film. It has been speculated that thickness can photocurrent. While we have not made a systematic analysis of
50 have a strong influence on the optical absorption because of this discrepancy, it is possible that the highly reflective
potential strain induced by the interaction of film and substrate.115 appearance of the film is connected with poor absorption of light,
Single crystalline hematite thin films obtained using the flux and that the film has a low porosity, small internal surface area
technique have been used for photoelectrochemical water 95 and thus a smaller solid-electrolyte interface area.
splitting with a photon collection efficiency of 20% at +0.5V.116 In a recent quantitative study, we have found that the
55 The positive bias was applied because there is insufficient band concentration of the precursor and weight of the dip-coated or
bending with no applied potential. These single crystalline spin-coated film influence the final photocurrent performance of
hematite electrodes were stable at pH 6.3 and -0.4 V. the films.117
100
 using these photoelectrodes, it was shown that two vicinal iron
atoms accumulated 4 holes during the reaction. Subsequently,
optimised Si-doped hematite films were prepared giving rise to a
current density of 4 mA cm–2 at 0.5 sun and a 2.1 % solar-to-
40 hydrogen conversion efficiency based on the hydrogen
combustion value of -280 kJ.mol-1.120 The photoanode
performance can be varied by the process parameters.121 For
example, the morphology can depend strongly on the temperature
and the silicon precursor. XPS studies confirmed that Si is
45 incorporated into the hematite lattice as Si4+ in APCVD-deposited
Si-doped films.122 The optical band gap increased with the
increasing TEOS flow rate as a result of decreasing size of the
nanocrystallite and quantum size confinement effects. From a
study conducted on reactive ballistic deposited Si-doped hematite
50 films, it was found that Si aggregated on the hematite surface,
passivating it and aiding charge transfer to the solution.123 Spray
pyrolysis using [Fe(acac)3] has also been used for the deposition
of Si-doped hematite films that exhibit a photocurrent density of
370µA cm–2 under 80mW cm–2 AM 1.5 solar illuminations with
55 0.2% Si doping.124
Other elements such as Zr, Cr, Mo, Zn, Cd, Ni, Pt and Ge have
been investigated as dopants for increasing the efficiency of
hematite films. Ti-doped films were fabricated by deposition-
annealing using titanium butoxide and iron(III) chloride as
60 precursors.125 The film had a photocurrent density of 1.83mA cm–
2
vs. RHE before titanium doping. The Ti-doping increased the
Figure 6. top - X-ray diffractogram of a hematite film dip-coated on FTO
photocurrent density up to 3 mA cm-2. The donor density also
glass with 500°C heat treatment for 30 minutes in air. Solid vertical lines
5 indicate bold Bragg indices from α-Fe2O3; dashed lines point to Bragg increased along with a consequent reduction in electron hole
peaks from the FTO substrate. Bottom - FESEM images of α-Fe2O3 recombination at a timescale beyond a few picoseconds. It was
grown by dip-coating on FTO glass, and current–voltage characteristics in 65 proposed that this led to the high photocurrent density. Sol-gel
the dark and under 1.5AM simulated light at a scan rate of 10 mV s–1. deposited Ti-doped films exhibit an IPCE of 32.6% at 400 nm.126
Dashed line at the bottom is dark current, dashed line at top is light
10 current, and solid thick line is the photocurrent (= light current – dark
By comparing hematite films doped with Si and Ti, it was
current). demonstrated that Si-doping led lower photocurrent densities than
Ti-doping.127 This is believed to be due to high grain boundary
70 recombination as well as a highly disordered surface with smaller
Doping-enhanced photoactivity of pristine hematite films size of the crystallites. The Ti-doped film showed good electrical
conductivity. In another study, co-doping of Ti and Si resulted in
Because of the low efficiency of pristine hematite films for water an IPCE of 34% at 365 nm.128 Spray pyrolysis has also been used
15 photo-oxidation, significant effort has been devoted towards the for Ti-doping of hematite, giving material with a photocurrent
development of doped hematite films and Si doping is an 75 density of 340 µA cm–2 at zero bias and 1.98 mA cm–2 at 0.5V vs.
effective strategy for increasing the efficiency. For instance, SCE.129 The system showed a good solar-to-hydrogen conversion
Kennedy118 prepared Si-doped hematite films by the pressing and efficiency of 1.38% at 0.5 V vs. SCE with a total hydrogen
sintering of hematite and SiO2 powders. Sintering was carried out collection of 2.44 ml per hour. A combinatorial technique has
20 at 1250ºC to 1350ºC for 4 hours followed by quenching at room also been utilized for the deposition of libraries of doped
temperature. The photocurrent varied consistently with the silicon 80 electrodes. Iron salts and Ti-, Si- or Al-doping precursors were
content and at 2 atom % Si the photocurrent density was 140 µA ink jet printed on conducting glass substrates using quantitative
cm-2. The photoresponse of the electrode was improved by the printing technique with subsequent annealing of the yielding a
treatment with 1 M KI solution, leading to an improvement in library of Ti-, Si- or Al-doped hematite photoanodes.130 The
25 photocurrent by a factor of three. This photocurrent was stable for electrodes were screened by immersing in electrolyte followed by
at least 2 months but disappeared on heating to 130 ºC. 85 laser scanning to measure the photocurrent. It was found that Ti-
The Grätzel group used atmospheric chemical vapour doping enhanced the photocurrent better than Si- and Al-doping.
deposition on SnO2-coated glass using Fe(CO)5 and TEOS In electrodeposited Zr-doped films, a photocurrent yield of 2.1
precursors to prepare Si-doped hematite photoelectrodes with an mA cm–2 at 0.6 V vs. SCE was obtained.131
30 IPCE of 42%.119 An HRSEM study demonstrated highly dendritic Moving to other transition metal dopants, enhanced
nanostructures of 500 nm with surface features of 10-20 nm. It 90 efficiencies for photoelectrochemical water splitting were
was suggested that the high photocurrent obtained (2.2 mA cm–2) observed with Cr- or Mo-doped films.132 At doping levels of 5
was because of the dendritic nanostructure which minimized the atom % Cr or 15 atom % Mo, IPCEs of 6% and 12% at 400 nm
hole diffusion length and optimises light absorption. From a and an applied potential of 0.4 V vs. Ag/AgCl were observed
35 mechanistic investigation of water photo oxidation chemistry
 with the increased efficiency attributed to enhanced charge found that below the water splitting potential, electrolyte
transfer. 60 oxidation limits the performance of hematite electrode whereas
Cadmium has been incorporated as dopant into hematite by co- above 1.23 V, bulk recombination played a leading role in
electrodeposition of aqueous Fe(III) and Cd(II) at varying Cd : Fe limiting the photo-oxidation process.
5 ratios.133 The 1 atom % Cd-doped films showed 4-fold increases In related studies, photoelectrode bilayers based on tungsten
in photocurrent density over pristine hematite. Electrodeposition oxide and hematite were prepared by sol-gel methods using PEG-
has also been used for the preparation of Ni-doped hematite films 65 300 polymer as a structure directing agent.144 To circumvent
with improved charge transport and increased recombination effects of poor light absorption and small hole diffusion length, a
lifetimes and the highest photocurrents obtained were 1.5 mA host guest approach was used in which a high surface area
10 cm–2 at 0.65 V vs. Ag/AgCl in 0.1 M NaOH.134 Pt-doped tungsten oxide host film made by APCVD supports a thin layer
hematite films have also been prepared by electrodeposition with of hematite.145 This forms photo-generated holes very close to the
an IPCE efficiency 4-fold higher than pristine hematite (12% vs. 70 semiconductor-electrolyte interface. The increase in photocurrent
3%) at 400 nm.135 The overall energy efficiency of the Pt-doped density was around 20% compared to a control film deposited
hematite was around 3% at 400 nm at 0.46 V in a two electrode without the host scaffold and the APCE was believed to be
15 system. Finally, Ge-doped hematite gave a short circuit current responsible for the enhanced photocurrent density. An 8%
of 1.2 mA cm–2 with a power conversion efficiency of 0.06% in improvement in APCE was observed as compared to 5.7% for
the presence of additional redox-active species.136 75 control films.
Surface treatment of hematite photoanodesIn addition to doping,
Hematite based 1D, 2D and 3D nanostructures for PEC
other strategies have been adopted for improving the photocurrent application
20 of hematite films and increasing the efficiency for water splitting.
One approach is the overlayer treatment of ultrathin hematite Hematite nanostructures with different dimensionality and levels
films with group 13 oxides. A Ga2O3 surface treatment lowered of complexity have been investigated for PEC applications. One
(= improved) the photocurrent the onset potential for water 80 recent review on nanostructured hematite appeared in this
oxidation by up to 200 mV.137 The photocurrent was sustainable journal. 146a The hopes behind nanostructuring are to increase the
25 for at least one day. Water splitting with zero bias has been efficiency of conventional particulate-based hematite films. This
observed for Ti-doped hematite films treated with an aqueous approach helps in optimizing factors such as the hole diffusion
solution of CoF3 to negative shift the flat band potential and length or high recombination rates.
allow direct reduction of water into hydrogen by the 85 Hematite nanowires have been synthesized by solution growth
photogenerated electrons.138 Addition of glucose by-passed the and post-annealing of hydrated fluoride nanowires.146b Optimized
30 slow oxygen evolution and provided electrons to consume the hematite nanowire electrodes gave a photocurrent density of 540
photo-generated holes. Ti-doped hematite films have also been µA cm–2 at 1.23 V vs. RHE. Due to the ease of synthesis, this
modified with electrodeposited NiFe binary oxides to boost the method can be used for large scale production of hematite based
water photooxidation reaction.139 The electrocatalyst deposited 90 nanowires. Hydrothermal methods have also been utilized for the
from Ni(II) and Fe(II) precursors increased the IPCE of the synthesis of Sn-doped hematite nanowires and nanocorals for use
35 photoelectrode by as much as a factor of 5. as photoanodes.147 As-synthesized nanowire films were sintered
Treatment of 12.5 nm thin hematite films with TEOS resulted at higher temperatures to provoke interdiffusion of Sn from the
in a drastic improvement of the photoresponse with a water FTO substrate into the hematite leading to Sn-doping. Sn-doped
splitting photocurrent of 630 µA cm–2 and an onset potential of 95 hematite nanocorals were obtained by adding tin(IV) chloride
1.1 V vs. RHE.140 No photoresponse was observed for TEOS during the hydrothermal deposition process. The Sn-doped
40 untreated FTO substrate and the increase in photocurrent was hematite nanowires sintered at 800ºC showed a photocurrent
attributed to the good crystallinity of the hematite film and density of 1.24 mA cm–2 at 1.23 V vs. RHE. In case of the
reduction of the number of carrier trapping sites. In order to nanocorals, a 1.5 times higher photocurrent density of 1.86 mA
decouple the morphological and functional modification, a simple 100 cm–2 was obtained under similar operating condition. Ultrafast
solution-based strategy has been developed in which electrodes optical absorption spectroscopy showed that Sn-doping and
45 are encapsulated in an SiO2 confinement scaffold before heat morphology did not influence electron-hole recombination in the
treatment.141 This approach led to an increase in photocurrent picosecond regime; XPS studies confirmed that Sn substituted an
density from 1.57 mA cm–2 to 2.34 mA cm–2 under standard Fe3+ site in the lattice. It is believed that the enhanced
illumination conditions in 1M NaOH. The improved quantum 105 photocurrents for Sn-doped nanowires and nanocorals are a result
efficiency with longer wavelength photons is due to the small size of both improved electrical conductivity and increased surface
50 of the crystallite. Modification of hematite with a co-catalyst such area. Ti- and Sn-doped nanowires with the crystallographically
as IrO2 resulted in an enhanced photocurrent density of 3 mA cm– preferred direction [110] were obtained by annealing the [211]
2
at an applied potential of +1.23 V vs. RHE.142 During this oriented FeOOH nanowire arrays obtained from solvothermal
investigation it was proposed that parallel tuning of both the 110 deposition on an FTO substrate.148 The materials showed good
catalysis and nanostructure could lead to hematite electrodes with photocurrent densities of 1.3 mA cm–2 under AM 1.5 illumination
55 a fully exploited potential for water splitting. In a different and had an effective pathway for charge transport which further
approach, the photocurrent densities of hematite films were increased photocurrent. The length of the nanowire can be
increased using hole scavengers such as H2O2,143 allowing for the controlled by adjusting the reaction temperature as well as
separation of surface and bulk recombination effects. It was 115 solution pH. At low pH of the precursor solution, longer
nanowires were obtained due to the acid catalyzed hydrolysis of
 acetonitrile. A recently reported 2D hematite nanostructure can be directly overlaid with the symbol of the wind rose in the
exhibits the highest reported incident photon to current lower left part and shows a surprisingly accurate match of both.
conversion efficiency of 46% at 400 nm, even higher than the The wind rose, such as in the NATO symbol, is built up from
benchmark hematite electrode presented recently.149a The wings with inclined facettes, and our hexapod has such inclined
5 necessary heteronanostructures were fabricated by depositing a 55 facettes, which is convincingly seen in the splitting of Bragg
thin layer of hematite on TiSi2 nanonets 149b by atomic layer reflections in the electron diffraction image in the lower right part
deposition techniques. The precursors used for the deposition of in Figure 8.
hematite films were Fe2(tBuO)6 and water. A water splitting
photocurrent of 1.6 mA cm–2 at 1.23 V vs. RHE was reported.
10 Thermal oxidation of iron foil in an oxygen rich atmosphere gave
rise to hematite nanorods which showed a photocurrent density of
0.9 mA cm–2 at 1.8 V vs. RHE.150 In an oxygen-lean
environment, hematite nanocorals were obtained with a best
photocurrent density of 2.6 mA cm–2. Hematite nanotubes
15 prepared by sonoelectrochemical anodization of an Fe metal foil
have also been investigated for solar water splitting.151 The
thickness of the resultant film was 5 - 7 nm with smooth arrays of
nanotube having lengths of 3-4 µm. The hematite nanotube film
showed a photocurrent density of 1.41 mA cm–2 with a 50%
20 contribution from the visible portion of the solar spectrum.
We have recently developed a method for increasing the
efficiency of a pristine hematite film by a factor of 2 by
hydrothermal after-treatment.152 The treatment yields a 3-D
flower-like nanostructure superimposed on a bed of hematite
25 nanorods. This modification process involves the hydrothermal
60 Figure 8. TEM micrographs showing stellate (upper left) matching the
treatment of a pristine dip-coated hematite film with amino acid well-known wind rose structure (left top and bottom), wings with inclined
such as L-arginine in the presence of iron (III) chloride. In a faces and their nanometer sized sub-units (top right). The inclined face
typical experiment, the hydrothermal treatment is carried out in a wind rose structure is confirmed by the splitting of Bragg reflections in
closed glass vial containing 5 mmol each of FeCl3.6H2O and L- the crystallographically indexed selected area electron diffractogram
65 (bottom right).
30 arginine at 95ºC for 48 hours. The result of the transformation of
the nanoparticulate morphology into a flower-like morphology is The photocurrent of the modified hematite film was studied in
presented in the SEM images in Figure 7. 1M KOH under bias and AM 1.5 solar light illumination. The
photocurrent density increased from 148 μA cm–2 to 218 μA cm–2
after 48 hour of hydrothermal treatment (Figure 9) with the
70 increase being attributed to the larger surface area of nanoflower,
as determined from BET surface area measurements.

Figure 7. Hydrothermal treatment of a pristine hematite film (see SEM

35 inset in Figure 6) causes substantial microstructure transformation (left)
with formation of stellate nanostructures (right). Photo

The optical assessment of the nanoflower structures indicates

that the nanoflower films have lower scattering coefficients than
the pristine film due to the roughening of the film in the
40 hydrothermal process. The band gap appears to be unaffected by
hydrothermal treatment and remained in the same range as
determined for pristine hematite (1.9-2.2 eV). X-ray
Figure 9. Photocurrent density of the pristine and nanoflower hematite
diffractometry revealed that the structure of the pristine film
film at 450 mV vs Ag/AgCl. The dark currents have been subtracted.
transformed to the flower-like morphology with preferential
45 orientation and a distinct texture. Close inspection of the
75 Bioconjugates
nanoflower film (Figure 8) with scanning electron microscopy
revealed that the nanoflowers are hexapods resembling 3- Conjugates of iron oxide nanoparticles with biological motifs
dimensional wind roses with an absolute orthogonality with are of significant interest in the biomedical sector. For instance,
branches faceted parallel to the long axis. The TEM micrograph to study the interaction of iron oxide nanoparticles with proteins,
50 of the hexapod (or nanoflower) in the upper left part of Figure 8 maghemite nanoparticles were conjugated to bovine serum
 albumin protein via a fatty acid binding domain.153 Numerous for water splitting applications.
reports concern the application of iron oxide bioconjugates for Recently, a hydrogen evolving hybrid photocatalyst was
targeted drug delivery, MRI contrast agents or for sensing fabricated by the non-covalent interaction of photosystem I with
applications.154-158 Integration of biological photosynthetic 50 Pt nanoparticles.163 The photocatalytic rate obtained for PSI / Pt
5 functionalities on the semiconductor photoelectrode surface could biohybrid system was 244 µmol H2/mg chlorophyll / hour using
give new ways to improve the performance of solar cells. It might cytochrome c6 as the soluble mediator and ascorbate as sacrificial
also be possible to mimic the photosynthetic machinery of nature electron donor. The EPR characterization of the biohybrid
or to demonstrate artificial photosynthesis processes.159 Systems complex has elucidated the electron transfer reaction between PSI
with a biomolecule modified photoanode and photocathode are 55 and Pt metallic nanoparticle.
10 called bio-photoelectrochemical cells, which we have naively
sketched in Figure 10.
Bioconjugated photoelectrode based on metal and metal oxide
substrate
In order to make the transition from bio-conjugated particulate
60 systems to films and assemblies, we need to look back to 1971
when Tributsch and Calvin pioneered a hybrid semiconductor-
biological electrode for the study of photoelectrochemical
reactions of excited chlorophyll molecules at the semiconductor
surface.164 As electrodes they took single crystals of wide band
65 gap ZnO and showed that the optically excited chlorophyll
injected electrons into the CB of ZnO, thus giving an anodic
photocurrent. The presence of various electron donors such as
hydroquinone or phenylhydrazine in the electrolyte helped the
chlorophyll molecules in mediating the transfer of electrons from
Figure 10. Schematic representation of bio-photoelectrochemcial cell. The 70 the levels of the reducing agents into the CB of the
photoanode on the left, such as hematite is functionalized with semiconductor which acts as electron acceptor. When an
15 phycocyanin. Ther photocathode on the right, such as copper oxide, is appropriate electrochemical gradient was provided, the electron
functionalized with a hydrogenase. A proton conducting NAFION® capture by the semiconductor electrode was irreversible. The
membrane warrants ionic conductivity and prevents mixing of evolving
excited chlorophyll at semiconductor showed similarity in
O2 and H2.
75 properties such as unidirectional charge transfer and efficient
To date, little effort has been devoted to the development of charge separation to chlorophyll in the photosynthetic reaction
20 bioconjugated photoanodes for solar energy applications. Our center. However this prototype system has no galvanic
focus is primarily on the development of electrodes conjugated separation.
with biological molecules giving biomimetic synthetic analogues The organic-inorganic photogalvanic effect was first observed
of biological light harvesting molecules, the active site of the 80 by Rabinowitch in 1940 in the thionine–iron system.165 The
water oxidation catalyst or the hydrogen evolution catalyst. photogalvanic effect of chlorophyll was studied by attachment to
a metal electrode such as Pt in series with a quinhydrone half-
25 Bioconjugated Hybrid Photocatalysts
cell.166 In this study it was found that on illumination, chlorophyll
In addition to the fabrication of bioconjugated semiconductor a aggregates underwent charge transfer with the subsequent
photoelectrodes, hybrid photocatalysts have been developed for 85 formation of a p-type semiconductor film. This gave rise to a
solar fuel production. Tran et al. recently provided an interesting photocurrent with a positive potential at the Pt-Chl electrode. A
overview of how to functionalize the surface of photocatalyst reverse current was observed when the light was turned off. This
30 particles with H2 evolution catalysts, water oxidation catalysts reversible photogalvanic reaction was only observed upon
and carbon dioxide reduction catalysts, and also synthetic irradiation. The effect was absent on the chlorophyll-free side of
molecular organic catalysts. 160 90 the electrode. The observed photocurrent at 700 nm matched well
As an example, a hybrid system was fabricated which with that of chlorophyll as evident from action spectrum.
consisted of TiO2 nanoparticles and enzymes such as [NiFeSe]- Recently, an artificial leaf device was fabricated based on the co-
35 hydrogenase with co-attached Ru based photosensitizers. This immobilization of a chlorophyll derivative and formate
system showed promising results in producing hydrogen gas in an dehydrogenase on silica gel thin films for reduction of carbon
aerobic environment.161,162 To establish the per site performance 95 dioxide to formic acid. 167
of the hydrogenase in producing hydrogen, a similar system has Not only photosynthetic components, but also living organism
been developed using [NiFeSe]-hydrogenase from and the entire photosynthetic machinery such as photosystem II
40 Desulfomicrobium baculatum (Db) attached to Ru-dye sensitized and photosystem I have been utilized for the preparation of
TiO2. For that study, triethanolamine was used as a sacrificial hybrid phtotoelectrodes. For this, animate blue green algae
electron donor and the system produced H2 with a turnover 100 (mastigocladus laminosus) were immobilized on the surface of a
frequency of 50 (mol H2) s-1 (mol total hydrogenase)-1 at pH 7 SnO2 electrode.168 Such a living electrode was believed to be
and 25 °C, even under the typical solar irradiation of a northern applicable as long-lived photoconverter for converting sunlight
45 European sky. The system shows promising stability even after into electrical energy.
extended exposure to air, making it promising as robust catalyst Recently, Reisner developed a hybrid photoanode for water
 171
photooxidation made of cyanobacterial photosystem II and a on a pristine hematite film is a new approach for increasing
mesoporous indium doped tin oxide electrode.169 Charge transfer the efficiency of the water photo-oxidation process. The light
at the interface occurred via two competing pathways to the metal 45 harvesting protein acts as a photon-energy relay system in the
oxide interface with the application of minimum bias +0.2V vs. cyanobacterial membrane 174 and in our study, the protein helped
5 NHE during red light (635 nm) illumination. The photocurrent at in the panchromatic harvesting of low energy photons in the red
pH 6.5 and +0.5V was 1.6 ± 0.3 μA cm−2. These strategies are part of the visible spectrum giving an absorption threshold at 620
based on the chemical modification of electrode surface by nm. We adopted a covalent functionalization strategy to increase
electrostatic adsorption. To overcome the stability issue of the 50 the stability of the conjugate in the harsh operational environment
biomolecule attached to the electrode, covalent functionalization of a PEC. The conjugation process involved hierarchical
10 strategy is preferred over electrostatic adsorption modification of the hematite surface with protein, followed by
agarose and finally covalent modification of the agarose with a
Bioconjugated hematite photoelectrodes
second layer of C-phycocyanin by CDI-cross-linking, as outlined
To the best of our knowledge, bioconjugation of hematite 55 in Figure 11.
photoelectrodes is relatively new in the solar energy sector,
though an early study on the interaction between hematite and
15 cytochrome c, was presented by Egglestone in 2006.170 Recently
we have developed a method for functionalizing the surface of
hematite films with the light harvesting membrane protein C-
phycocyanin.171 The primary goal of developing such conjugates 60

is to increase the light-harvesting efficiency of hematite or

20 alternative semiconductors. Eggleston studied the interaction of
mitochondrial cytochrome c with hematite, which is important in
studying the mechanism of iron reduction by bacteria. The
cytochrome c was bound electrostatically to the hematite surface
in the pH range 8.5 - 10. Cyclic voltammetric studies confirmed
25 the presence of cytochrome c in its native conformation and the 65
protein adhered strongly to the hematite surface in the presence of
phosphate buffer as confirmed by deformation of the AFM tip
during atomic force microcopy scans.170 In related studies, the
protein was co-adsorbed on the hematite in phosphate buffer in
30 the pH range from 2 to 11. The protein aggregated with a Figure 12. Large scan AFM images (50 µm x 50 µm) of C-phycocyanin
conjugated film with fiber-like pattern on the hematite surface.
conformational change when the pH approaches the isoelectric
point of cytochrome c.172 The interaction of hydroquinone with 70 The bioconjugated film was characterized by a wide-range of
hematite surfaces has also been studied in order to elucidate the techniques. Optical spectroscopy revealed that the band gap of
mechanism of charge transfer.173 The catalytic oxidation of the original film of 1.9 - 2.2 eV was unchanged after the
35 hydroquinone on hematite involves an inner sphere 1-electron- hierarchical surface modification. Within the resolution and
transfer mechanism supported by the detection of semiquinone sensitivity limits of XRD, we found no structural changes in the
radicals. 75 hematite after chemical modification. The morphology of the film
was studied with tapping mode atomic force microscopy
revealing a fibre-like pattern on the surface after bioconjugation
(Figure 12). Raman spectromicroscopy exhibited a band at 1610
cm-1 confirming the presence of C-phycocyanin.

Figure11. Hierarchical functionalization of hematite with electrostatically

40 adsorbed and covalently-linked C-phycocyanin light-harvesting units.
Hematite-Phycocyanin conjugation scheme.
80
Bioconjugation of the light harvesting protein C-phycocyanin (a)
 (B)

Figure 13. a) The photocurrent for pristine hematite (solid line), C-

phycocyanin electrostatically adsorbed on hematite (dashed line), C-
5 phycocyanin conjugated (dotted line) films in 1M KOH (pH=13.6) with a
three electrode PEC cell at 600 mV bias under dark and simulated AM Figure 14. (a) Energy band diagram showing the charge transfer from
1.5 solar illumination. b) ) top- Photocurrent action spectrum of pristine 25 excited state phycocyanin to conduction band of hematite (b) The
hematite (solid line), C-phycocyanin electrostatically adsorbed on photocurrent transient responses of pristine hematite (solid line), C-
hematite (dashed line), C-phycocyanin conjugated (dotted line) films in phycocyanin electrostatically adsorbed on hematite (dashed), C-
10 1M KOH (pH=13.6) under monochromatic irradiation and with 0 mV phycocyanin conjugated (dotted line) films measured during
bias. Bottom - IPCE of pristine hematite film (solid), phycocyanin chronoamperometric cycling.
absorbed hematite film (dashed), and conjugated film (dotted line),
obtained with 0 mV bias. 30 The increase in photocurrent by a factor of two was attributed
to the enhanced panchromatic light harvesting of hematite film by
The enhanced efficiency of the bioconjugated film is evident the phycocyanin film which absorbs at 620 nm. IPCE
15 from a photoelectrochemical study using a sweeping current- measurements were carried out with 0 mV, i.e. no bias, and
potential scan in 1M KOH in dark and under AM 1.5 simulated further validated the enhanced efficiency of the bioconjugated
solar irradiation. The photocurrent density at 600 mV increases 35 system at 620 nm (Figure 13b). We have obtained 0.2% IPCE for
from around 288 µA cm–2 for the pristine hematite to around 480 pristine and protein adsorbed film which increases to 0.6% upon
µA cm–2 for the protein adsorbed film and to around 570 µA cm– conjugating the protein by surface functionalization concomitant
2
20 for the bioconjugated film (Figure 13a). Only the photocurrents with the increase in photocurrent, as is evident by the
are shown as the dark current densities have been subtracted. photocurrent action spectrum (upper, Figure 13b). The very low
40 value of IPCE originated as a result of the application of zero
bias.

C-Phycocyanin itself can also oxidize the water with the

generation of hydroxyl radicals 175 and we speculated that the
45 protein might undergo additional reactions with the hematite
surface or electrolyte under irradiation in the
photoelectrochemical study.
In the energy level diagram in Figure 14(a), we want to present
the charge transfer of electron into conduction band of hematite
50 similar to DSSC process. The valance band of the phycocyanin is
more negative than the water oxidation reaction, which will
prevent hole transfer from protein to water. The schematic is
insofar questionable. From literature, it is known that
phycocyanin acts as a photo-oxidizer due to its photodynamic
55 action, while the exact mechanism of hole transfer is still not
established. It is stated that phycocyanin can interact with water
 molecules (solvent) or with hematite (substrate) by a so-called
type-I mechanism, whereby it transfers charge to the solvent
(electrolyte in this case) with the generation of radicals. This
happens only when phycocyanin gets excited with light.
5 Photosensitized oxidations are the basis for photodynamic action.
The first step of this photodynamic action is the absorption of
light by a sensitizer (sens) to produce an excited state (sens*). In
the presence of oxygen, two competing reactions of the excited-
state sensitizer can occur, which are, respectively, Type-I and
10 Type-II mechanisms. Hydroxyl radicals (OH • ) can be generated
by the following mechanism when C-phycocyanin is irradiated
with visible light (470 nm). It is speculated that in our case, the
sens* can either react with the hematite surface or with the
electrolyte during photoelectrochemical operation (Type I) or
15 with oxygen (Type II).
To better understand the charge transfer between excited state C-
phycocyanin and the conduction band of hematite as depicted in
energy level diagram (see Figure 14 a), we carried out a transient
photocurrent study of the bioconjugated films (Figure 14 b). The
20 falling transient obtained during the study suggested the
photogeneration of a hydroxyl radical which can act as an
electron acceptor.15 The putative hydroxyl radical is depleted on
the hematite surface when the light is tuned off and scavenges the
CB electrons forming a cathodic transient spike. The spike decays
25 upon depletion of hydroxyl radicals. We observe that the
magnitude of this cathodic transient spike was maximum in case 50
of phycocyanin conjugated hematite film which lead us to Figure 15. a) Stability of photocurrent over time for pristine and protein
suggest that bioconjugated films generate more hydroxyl radical functionalized film. b) Number of moles of electrons obtained during dark
due to photodynamic action of phycocyanin molecule. We chronoamperometric scan after 10 min of operation for the three different
30 believe that this is probably another factor for enhanced treated samples.
efficiency. But this needs to verify in dedicated future studies. 55

To validate the effect of oxidation on the enhanced efficiency Influence of protein conjugation on optical absorption
of the bioconjugated films, we have performed a long term
35 chronoamperometry study at the oxidation potential of water (207 Light-harvesting action of the protein molecules which absorb in
mV vs. Ag/AgCl and 1.23 V vs. a reversible hydrogen electrode the red part of the visible spectrum helps in the anticipated
(RHE)] as shown in Figure 15 a. The primary origin of the 60 enhanced photocurrent obtained in hematite–phycocyanin system.
photocurrent is water oxidation by photogenerated holes, but with By integrating this protein on the hematite or photoactive layer,
the bioconjugated film, the photocurrent can also involve the we can increase its light-absorbing properties by taking advantage
40 faradaic oxidation of protein. The chronoamperometry study panchromatic light harvesting. Multiple factors affect the total
indicated that the photocurrent was stable for bioconjugated number of photons absorbed by a solar cell active layer. These
hematite. These results lead us to believe that the extra current is 65 are: an optical absorption comparable to the solar spectrum, the
not due to irreversible oxidation of protein. Figure 15 b. displays absorptivity, and the broadening of the absorption-wavelength
the number of moles of electrons generated under dark condition range. In this system, the generation of a photocurrent or
45 and as is evident, the concentration of electrons generated during efficiency will be enhanced only when the total number of
dark oxidative process decreases for bioconjugated samples. This photons absorbed by it increases. This could be possible once the
supports that the observed extra current is not from the oxidation 70 absorption range and absorption peak value or the intensity
of the organic molecules on the hematite surface. increases for hematite. So, by anchoring the C-phycocyanin
molecule on the hematite surface, the absorption peak value can
be increased with a broadening of the spectral range. The
absorption spectra shown in Figure 3a in ref. 171 reveal that the
75 pristine hematite film absorbed at 404 nm, 444 nm and 545 nm,
corresponding to the usual band positions of bulk hematite. The
band at 444 nm originated as a result of ligand field transition and
that at 545 nm appeared due to absorption edge. Protein
adsorption helps in spectral broadening along with a shift of
80 444nm band to longer wavelength, while no change to the band at
404 nm. It is found that absorption edge peak intensity increases
upon protein interaction. This change can be related to the change
 of the spectral weight and the increase in the peak absorption transfers between semiconductor and protein moiety, we are still
value occurs due to protein absorption edge at 620nm. C- 60 away from a conclusive statement about the potential removal of
phycocyanin absorbs at 620 nm, and this can be seen from the surface states by phycocyanin. Our hybrid integrated system can
inset in Figure 3a in ref. 171. The absorption maximum of work as a future DSSC system fabricated with affordable and
5 phycocyanin changes upon interaction with the hematite environmentally benign components such as hematite and
nanoparticles constituted the film. It was due to the perturbation phycocyanain. In this case, we need to study the regeneration of
of the surface-charge density of the nanoparticle and the protein, 65 chromophore of phycocyanin molecule which is an important
which manifests in changing the absorption spectrum. issue in DSSC application. It is noteworthy that phycocyanain
production is today an easily scalable process. In our system, we
Hematite conjugates with synthetic biomimetic analogues
can enhance the efficiency in two approaches: adsorption and
10 The attachment and covalent linking of redox-active and photo- chemical cross functionalization. We have found that chemical
active molecular motifs to semiconductors lies at the heart of the 70 cross functionalization stabilized the light harvester.
dye-sensitized solar cell (DSSC).176 Typically, semiconductors
such as TiO2 or ZnO are associated with the concept of DSSC Conclusions and Outlook
applications, but some examples have been reported in which
hematite films have been covalently modified with synthetic We have discussed the chemical and physical characteristics of
15

molecules. hematite and its utility as a photoanode for the PEC water
Is it possible, or even necessary to improve the photo response splitting reaction. The focus is on how various synthetic
of hematite by dye-sensitization permitting the harvesting of 75 approaches can be used to modify the properties of hematite
incident photons with higher efficiency? For example, nanoparticles and thin films to increasing the efficiency for solar
20 electrodeposited hematite films have been further functionalized water splitting. Approaches to PEC including microstructure
by physisorption of [(5, 10, 15, 20-tetrakis(4-carboxyphenyl) refinement and bioconjugation of hematite with photosynthetic
porphyrin (TCPP)] and the efficiency of the TCPP modified counterparts have been reviewed. We believe that these
electrode was 70% higher compared to the bare hematite 80 approaches may be generally applicable for further increasing the
electrode.177 efficiency of well-performing PEC materials.
25 Dip-coating of hematite film and surface functionalization for
The dye sensitization of hematite is a fairly new field. In the case bioconjugation, paves new ways towards cost effective PEC cells
of bare hematite, major current research efforts focus on the light for H2 production.
absorbing properties, charge carrier generation and recombination 85 The two fold increase in photocurrent observed on
losses. However, the performance is also dependent on the charge bioconjugation is effective when we consider large area electrode
30 transfer and collection of light generated electrical charge carrier. fabrication, whereas the production of the prototype protein of
This latter property can be boosted with the help of external choice, C-Phycocyanin could pose a cost issue which needs to be
charge injection from a sensitizer into the conduction band of the technologically addressed as well.
semiconductor. Non-doped hematite has low efficiency because 90 One can consider extending the strategy with other
of the short average lifetime of the excitonic pair, i.e. the short photosynthetic molecular units or dyes. A significant factor is the
35 hole diffusion length. 178 This means most of the holes recombine stability of these modified electrodes in the harsh PEC operation
before they can diffuse to the surface and do the water oxidation environment.
reaction, and consequently the efficiency becomes low. Here, One general heard concern is that biomaterials such as proteins
surface states or Fe2+ ions from less-oxidized phases such as 95 are too sensitive for any practical application. This may hold if
FeOOH at the surface 179 can play an important role on the one keeps those motifs, or even dyes from standard DSSC, in a
40 recombination process. To cover up these surface states various liquid environment for analytical purposes. However, once
strategies have been followed. Dye sensitization is one alternative attached and immobilized on surfaces, such seemingly sensitive
technique in the sense that the OH group from the surface state materials may show more robustness than we actually expect. So
provides a room for the formation of molecular species which act 100 there is hope that degradation issues may have been exaggerated,
as bridge for charge transfer across semiconductor-molecule and that the handling of biological or biochemical components
45 interfaces. Then, when we install a sensitizer on top of hematite, could become easier at the device level.
the HOMO of which lying higher above the VB of hematite, we Finally, we are confident that hematite films obtained by dip
can increase the efficiency to some extent. The sensitizer acts so- coating, further refined by hydrothermal treatment and
to-speak as a panchromatic light harvester which enhances the 105 functionalized with phycocyanin can be eventually produced as
spectral selectivity and catalytic action of the electrode. large PEC photoanode panels at an industrial scale.
50 Porphyrin is such sensitizer mimicking the structure of the natural
chlorophyll molecule. Verrara et. al. 177 used a derivative of Acknowledgements
porphyrin for the dye sensitization of electrodeposited hematite
film. The stability of biological materials like dyes and proteins is This research was fully funded by the Swiss Federal Office of
in general under scrutiny when they are considered in ambient Energy project No. 153613-102809 and partially funded by the
55 envionment. From the stability perspective, we find that C– 110 Swiss National Science Foundation project R’Equip # 121306.
phycocyanin can work in harsh photoelectrochemical condition, Electron micrographs were provided by Empa Electron
such as in pH ~ 14 under 1.5 AM illumination and strong anodic Microscopy Center (Dr. Rolf Erni) and the Argonne National
bias. While we have investigated the mechanism of charge Laboratory Center for Nanoscale Materials (Dr. Elena Rozhkova
 and Dr. E. Vitol). X-ray and electron spectroscopy data were 55 17. M. P. Dare-Edwards, J. P. Goodenough, A. Hamnett and
obtained at the Advanced Light Source in Berkeley and at P. R. Trevellick, J. Chem. Soc., Faraday Trans., 1983, 79,
BESSY-II Berlin and at ESRF Grenoble. We are grateful to Dr. 2027.
Krisztina Schrantz (funded by the Swiss-Hungarian Sciex project 18. (a) L. Pauling and S. B. Hendricks, J. Am. Chem.
5 10.013) for assistance with the bio conjugation, and to MSc Soc., 1925, 47 , 781; (b) R. M. Cornel and U.
student Pradeep Wyss (funded by the Hirschmann Foundation) 60 Schwertmann, The Iron Oxides. Structure, Properties,
and Prof. H. Piles, FHNW Basel. Graphic enhancement of the Reactions and Uses, VCH, Weinheim, Germany, 1996.
SEM images was carried out by A. Koster (Empa). 19. (a) L. A. Marusak, R. Messier and W. B. White, J. Phys.
Chem. Solids, 1980, 41, 981; (b) T. B. Saito, Bull. Chem.
Notes and references Soc. Jpn. 1965 , 38 , 2008; (C) L. Tay, N. L. Rowell, D. J.
10
a
Laboratory for High Performance Ceramics, Empa. Swiss Federal 65 Lockwood and R. Boukherroub, J. Vac. Sci. Technol. A
Laboratories for Materials Science and Technology, CH-8600 2006, 24, 747.
Dübendorf, Switzerland. Fax: +41 58 765 4150; Tel: +41 58 765 4850; 20. J. H. Kennedy and K. W. Frese, J. Electrochem. Soc.
E-mail: [email protected]
b 1978, 125, 709.
Department of Chemistry, University of Basel, CH-4056 Basel,
15 Switzerland. Fax: +41 61 267 1020; Tel: +41 61 267 1001; E-mail: 21. Y. P. He, Y. M. Miao, C. R. Li, S. Q. Wang, L. Cao, S. S.
[email protected] 70 Xie, G. Z. Yang and B. S. Zou, Phys. Rev. B 2005, 71,
c
Advanced Light Source, Lawrence Berkeley National Laboratory, 125411 d.
Berkeley CA 94720, USA. Fax: +1510 486 7588; Tel: +1 510 486 7831;
22. Y. Yamanoi, S. Nakashima and M. Katsura, Am. Mineral.
E-mail: [email protected]
2009, 94, 90.
23. (a) S. V. Yanina and K. M. Rosso, Science 2008, 320,
20 1. H. Gaffron and J. Rubin, J. Plant Physiol., 1942, 26, 219.
75 218. (b) B. Klahr, S. Gimenez, F. Fabregat-Santiago, J.
2. http://www1.eere.energy.gov/hydrogenandfuelcells/tech_
Bisquert and T. Hamann, Energy & environmental
validation/pdfs/fcm01r0.pdf
Science 2012, 5, 7626.
3. M. G. Walter, E. L. Warren, J. R. McKone, S. W.
24. Z. Y. Wu, S. Gota, F. Jollet, M. Pollak, M. Gautier-Soyer
Boettcher, Q. Mi, E. A. Santori and Nathan S. Lewis,
and C. R. Natoli, Phys. Rev. B, 1997, 55, 2570.
25 Chem. Rev. 2010, 110, 6446.
80 25. Y. Ma, P. D. Johnson, N. Wassdahl, J.H. Guo, P. Skytt, J.
4. W. Lubitz, E. J. Reijerse and J. Messinger, Energy
Nordgren, S. D. Kevan, J. E. Rubensson, T. Beoske and
Environ. Sci., 2008, 1, 15.
W. Eberhardt, Phys. Rev. B, 1993, 48, 2109.
5. H. J. Lewerenz, Chem. Soc. Rev., 1997, 26, 239.
26. E. C. Todd, D. M. Sherman and J. A. Purton, Geochim.
6. E. Becquerel, C. R. Acad. Sci. 1839, 9, 561.
Cosmochim. Ac., 2003, 67, 881.
30 7. H. Gerischer, J. Electroanal. Chem. Interf. Chem., 1975,
85 27. Z. Y. Wu, S. Gota, F. Jollet, M. Pollak, M. Gautier-Soyer
58, 263.
and C. R. Natoli, Phys. Rev. B, 1997, 55, 2570.
8. A. Fujishima and K. Honda, Nature, 1972, 238, 37.
28. B. Gilbert, C. Frandsen, E. R. Maxey and D. M. Sherman,
9. M. Calvin, Science, 1974, 184, 375.
Phys. Rev., 2009, 79, 035108.
10. A. J. Nozik, Ann. Rev. Phy. Chem.1978, 29, 189.
29. S. U. M. Khan and J. Akikusa, J. Phys. Chem. B, 1999,
35 11. (a) H. H. Kung, H. S. Jarrett, A. W. Sleight, and A.
90 103, 7184.
Ferretti, J. Appl. Phys., 1977, 48, 2463; (b) Z. Chen, T. F.
30. L. Vayssieres, C. Sathe, S. M. Butorin, D. K. Shuh, J.
Jaramillo, T. G. Deutsch, A. Kleiman-Shwarsctein, A. J.
Nordgren, J. H. Guo, Adv. Mater. 2005, 17 (19), 2320.
Forman, N. Gaillard, R. Garland, K. Takanabe, C. Heske,
31. (a) D. K. Bora, A. Braun, S. Erat, A. K. Ariffin, R.
M. Sunkara, E.W. McFarland, K. Domen, E. L. Miller, J.
Lohnert, K. Sivula, J. Topfer, M. Gratzel, R. Manzke,T.
40 A. Turner and H. N. Dinh, J. Mater. Res., 2010, 25, 3 ; (c)
95 Graule and E. C. Constable, J. Phys. Chem. C, 2011, 115,
F.E. Osterloh, Chem. Mater. 2008, 20, 35–54. (d) J. E.
5619. (b) A. Braun , K. Sivula, D. K. Bora, J. Zhu, L.
Katz, X. Zhang, K. Attenkofer, K. W. Chapman, C.
Zhang, M. Gratzel, J. Guo and E. C. Constable J. Phys.
Frandsen , P. Zarzycki, K. M. Rosso, R. W. Falcone, G.
Chem. C, 2012, 116, 16870.
A. Waychunas and B. Gilbert, Science, 2012, 337, 1200.
32. (a) S. Mochizuki, Phys. Stat. Sol. 1977, 41, 591; (b) C.
45 12. M. Yellishetty, P.G. Ranjith and A. Tharumarajah,
100 Gleitzer, J. Nowotny und M. Rekas, Appl. Phys. A Mater.,
Resour. Conserv. Recy., 2010, 54, 1084.
1991, 19, 310.
13. S. Zhu, N. Huang , L. Xu, Y. Zhang, H. Liu, H. Sun and
33. A. J. Boman and H. J. Van Daal, Adv. Phys., 1970, 19, 2.
Y. Leng, Mater. Sci. Eng. C, 2009, 29, 1589.
34. F. G. Gardner, F. Sweet and D. W. Tanner, J. Phys.
14. C. J. Sartoretti, B. D. Alexander, R. Solarska, I. A.
Chem. Solids, 1963, 24, 1183.
50 Rutokowska, J. Augustynski and R. Cerny, J. Phys.
105 35. R. N. Bhowmik and A. Saravanan, J. Appl. Phys., 2010,
Chem. B, 2005, 109, 13685.
107, 053916.
15. K. L. Hardee, A. J. Bard, J. Electrochem. Soc., 1976, 123,
36. F. Bødker and S. Mørup, Europhys. Lett. 2000, 52, 217.
1024.
37. J. H. Kennedy and K. W. Frese, J. Electrochem. Soc.
16. W. Gärtner, Phys. Rev., 1959, 116, 84.
1978, 125, 709.
110 38. R. K. Quinn, R. D. Nasby and R. J. Baughman,
Mater.Res.Bull.1976, 11, 1011.
 39. C. Sanchez, K. D. Sieber and G.A. Somorjai, J. 64. C. L. Snow, C. R. Lee, Q. Shi, J. Boerio-Goates and B. F.
Electroanal. Chem., 1988, 252, 269. Woodfield, J. Chem. Thermodyn., 2010, 42, 1142.
40. A. A. Tahir, K. G. Upul Wijayantha, S. Saremi- 65. C. Carbone, F. Di Benedetto, C. Sangregorio, P.
Yarahmadi, M. Mazhar, and V. McKee, Chem. Mater., 60 Marescotti, L.A. Pardi and L. Sorace, J. Phys. Chem. C,
5 2009, 21, 3763. 2008, 112, 9988.
41. A. Heller, K.C. Chang and B. Miller, J. Electrochem. 66. C.B. de Boer, T. A. T. Mullender and M. J. Dekkers,
Soc., 1977, 124, 697. Geophys. J. Int., 2001, 146, 201.
42. A. Heller, K.C. Chang and B. Miller, J. Am. Chem. Soc., 67. S. H. Gee, Y. K. Hong, J. C. Sur, D. W. Erickson, M. H.
1978, 100, 684. 65 Park and F. Jeffers, IEEE Trans. Magn., 2004, 40, 2691.
10 43. L.M. Peter, K. G. Upul Wijayantha and A. A. Tahir, 68. A. H. Hill, F. Jiao, P. G. Bruce, A. Harrison, W.
Faraday Discuss., 2012, 155, 309. Kockelmann and C. Ritter, Chem. Mater., 2008, 20, 4891.
44. Q. Zhang and G. J. Adriaenssens, J. Appl. Phys. 1997, 81, 69. C. Cromphaut, V.G. Resende, E. Grave and R.E.
6795. Vandenberghe, Hyperfine Interact., 2009, 191, 167.
45. M. Barroso, C. A. Mesa, S. R. Pendlebury, A. J. 70 70. G. F. Goya, M. Veith, R. Rapalavicuite, H. Shen and S.
15 Cowan, T. Hisatomi, K. Sivula, M. Grätzel, D. R. Mathur, Appl. Phys. A: Mater.Sci. Process. 2005, 80,
Klug, and J. R. Durrant, P. Natl. Acad. Sci.USA, 2012 1523.
[In Press] 71. S. N. Klausen, K. Lefmann, P. A. Lindgard, L. T. Kuhn,
46. P. H. Borse, H. Jun, S. H. Choi, S. J. Hong and J. S. Leeb, C. R. H. Bahl, C. Frandsen, S. Moerup, B. Roessli, N.
Appl. Phys. Lett. 2008, 93, 173103. 75 Cavadini and C. Niedermayer, Phys. Rev. B: Condens.
20 47. T. H. Jeon, W. Choi and H. Park, J. Phys. Chem. C 2011, Matter Mater. Phys., 2004, 70, 214411.
115, 7134. 72. M. V. Mansilla, R. Zysler, D. Fiorani and L. Suber,
48. T. K. Townsend, E. M. Sabio, N. D. Browning and F. E. Physica B, 2002, 320, 206.
Osterloh, Energy Environ. Sci., 2011, 4, 4270. 73. R. D. Zysler, D. Fiorani, A. M. Testa, M. Godinho, E.
49. L. Chen, G. Wang, G. N. Mathur, and V. K. Varadan, 80 Agostinelli and L. Suber, J. Magn. Magn. Mater. 2004,
25 Proc. SPIE, 2012, 8344, 83441J. 272, 1575.
50. B. Sun, J. Horvat, H. S. Kim, W. Kim, J. Ahn and G. 74. R. D. Zysler, D. Fiorani, A. M. Testa, L. Suber, E.
Wang, J. Phys. Chem. C, 2010, 114, 18753. Agostinelli and M. Godinho, Phys. Rev. B: Condens.
51. R. Prucek, M. Hermanek and R. Zboril, Appl. Catal. A, Matter Mater. Phys., 2003, 68, 212408.
2009, 366, 325. 85 75. U. M. Khan and J. Akikusa, J. Phys. Chem. B, 1999, 103,
30 52. D. Y. Wang, Z. Q. Liu, F. G. Liu, X. T. Zhang, Y. Cao, J. 7184.
F. Yu, T. H. Wu, Y. B. Bai, T. J. Li and X. Y. Tang, 76. A. Fujimori, M. Saeki, N. Kimizuka, M. Taniguchi and S.
React. Kinet. Catal. Lett., 1998, 65, 233. Suga, Phys. Rev. B, 1986, 34, 7318.
53. W. Zhang, J. Chen, X. Wang, H. L. Qi and K. S. Peng, 77. D. M. Sherman, Phys. Chem. Miner., 1985, 12, 161.
Appl. Organomet. Chem., 2009, 23, 200. 90 78. D.M. Sherman, Geochim. Cosmochim. Acta, 2005, 69,
35 54. F. Herrera, A. Lopez, G. Mascolo, P. Albers and J. Kiwi, 3249.
Appl. Catal. B: Environ. 2001, 29, 147. 79. B. Gilbert, C. Frandsen, E. R. Maxey and D. M. Sherman,
55. Y. Wang, W. Du and Y. Xu, Langmuir 2009, 25, 2895. Phys. Rev. B: Condens. Matter Mater. Phys. 2009, 79,
56. K. Sivula, R. Zboril, F.L. Formal, R. Robert, A. 035108.
Weidenkaff, J. Tucek, J. Frydrych and M. Grätzel, J. Am. 95 80. F. M. F. de Groot, M. Grioni, J. C. Fuggle, J. Ghijsen, G.
40 Chem. Soc. 2010, 132, 7436. A. Sawatzky and H. Petersen, Phys. Rev. B 1989, 40,
57. J. P. Crocombette, M. Pollak, F. Jollet, N. Thromat and 5715..
M. Gautier-Soyer, Phys. Rev. B. 1995, 52, 3143. 81. Z. Y. Wu, S. Gota, F. Jollet, M. Pollak and M. Gautier-
58. D. K. Bora, A. Braun, S. Erat, O. Safonova, T. Graule and Soyer, Phys. Rev. B, 1997, 55, 2570.
E. C. Constable, Curr. Appl. Phys. 2012, 12, 817. 100 82. W. C. Mackrodt, F. Joliet and M. Gautier-Soyer, Philos.
45 59. I. V. Chernyshova, M. F. Hochella Jr and A. S. Madden, Mag. B,1999, 79, 25.
Phys. Chem. Chem. Phys., 2007, 9, 1736. 83. E. Guiot, Z. Y. Wu, S. Gota and M. Gautier-Soyer, J.
60. R.V. Morris, H.V. Lauer Jr., C. A. Lawson, E. K. Gibson Electron Spectrosc. Relat. Phenom., 1999, 371, 101.
Jr., G. A. Nace and C. Stewart, J. Geophys. Res. B, 1985, 84. C. F. Gilbert, E. R. Maxey and D. M. Sherman, Phys. Rev.
90, 3126. 105 B, 2009, 79, 035108
50 61. A. Navrotsky, L. Mazeina and J. Majzlan, Science, 2008, 85. S. Dutta, S. K. Manik, M. Pal, S. K. Pradhan, P. Brahma
319, 1635. and D. Chakravorty, J. Magn. Magn. Mater., 2005, 288,
62. S. Morup, C. Frandsen, F. Bodker, S. N. Klausen, K. 301.
Lefmann, P. A. Lindgard and M. F. Hansen, Hyperfine 86. L.C. Sanchez, J. D. Arboleda, C. Saragovi, R.D. Zysler
Interact., 2003, 144, 347. 110 and C.A. Barrero, Physica B , 2007, 389, 145.
55 63. D. Spagnoli, B. Gilbert, G. A. Waychunas and J. F. 87. R. A. Borzi, S. J. Stewart, G. Punte, R. C. Mercader, M.
Banfield, Geochim. Cosmochim. Acta, 2009, 73, 4023. Vasquez-Mansilla, R. D. Zysler and E. D. Cabanillas, J.
Magn. Magn. Mater. 1999, 205, 234.
 88. R. Arbain, M. Othman and S. Palaniandy, Miner. Eng., 50 111. N. T. Hahn, H. Ye, D. W. Flaherty, A. J. Bard and C.
2011, 24, 1. Buddie Mullins, ACS Nano, 2010 , 4 , 1977.
89. M. A. Willard, L. K. Kurihara, E. E. Carpenter, S. Calvin
and V. G. Harris, Int. Mater.Rev., 2004, 49, 125. 112. C. S. Enache, Y. Q. Liang and R. Van de Krol, Thin Solid
5 90. Č. Jovalekić, M. Zdujić, A. Radaković and M. Mitrić, Films, 2011, 520, 1034.
Mater. Lett. 1995, 24, 365.
91. W. Wang, J. Y. Howe and B. Gu, J. Phys. Chem. C, 113. F. L. Souza , K. P. Lopes , P. A. P. Nascente and E. R.
2008, 112 , 9203. 55 Leite, Sol. Energ. Mat. Sol. C, 2009, 93, 362.
92. E. Darezereshki, Mater. Lett. 2011, 65, 642. 114. Y. Chen and K. Tu, Int. J. Photoenergy , 2011, 2012,
10 93. H.Tamura and E. Matijevic, J. Colloid Interf. Sci., 1982, 980595.
90, 100. 115. F. L. Souza, K. P. Lopes, E. Longo and E. R. Leite, Phys.
94. P. Tartaj , M. D. Puerto Morales ,S. Veintemillas- Chem. Chem. Phys., 2009, 11, 1215.
Verdaguer, T. Gonzá lez-Carreñ o and C. J. Serna, J. Phys. 60 116. R. K. Quinn, R. D. Nasby and R. J. Baughman, Mat. Res.
D: Appl. Phys. 2003, 36, R182. Bull., 1976, 11, 1011.
15 95. J. Rockenberger, E. C. Scher and A. P. Alivisatos, J. Am. 117. Bachelor Thesis, Romy Löhnart, Synthesis and
Chem. Soc., 1999, 121, 11595. characterization of hematite nanoparticles and films for
96. T. Hyeon, S. S. Lee, J. Park, Y .Chung and H. B. Na, J. photoelectrochemical applications, 2010, University of
Am. Chem. Soc., 2001, 123, 12798. 65 Applied Sciences, Jena, Germany
97. S. Sun, H. Zeng, D. B. Robinson, S. Raoux, P. M. Rice, S. 118. J. H. Kennedy, R. Shinar and J. P. Ziegler, J.
20 X. Wang and G. Li, J. Am. Chem. Soc., 2004, 126, 273. Electrochem. Soc., 1980, 127, 2307.
98. G. Zou, K. Xiong, C. Jiang, H. Li, Y. Wang, S. Zhang 119. A. Kay, I. Cesar and M. Grätzel, J. Am. Chem. Soc.,
and Y. Qian, Nanotechnology, 2005, 16,1584. 2006, 128, 15714.
99. Z. Jing and S. Wu, Mater. Lett., 2004, 58, 3637. 70 120. I. Cesar, A. Kay, J. A. Gonzalez Martinez and M. Gratzel,
100. M. Vasquez Mansilla, R. Zysler, D. Fiorani and L. Suber, J. Am. Chem. Soc., 2006, 128, 4582.
25 Physica B: Condensed Matter, 2002, 320, 206. 121. I. Cesar, K. Sivula, A. Kay, R. Zboril, and M. Gratzel, J.
Phys. Chem. C , 2009, 113, 772–782
101. R. Strobel and S. E. Pratsinis, Adv. Powder Technol., 122. S. Saremi-Yarahmadi, K. G. Upul Wijayantha, A. A.
2009, 20, 190. 75 Tahir and B. Vaidhyanathan, J. Phys. Chem. C, 2009,
102. H. Liu, Y. Wei, P. Li, Y. Zhang and Y. Sun, Mater. 113, 4768.
Chem. Phys., 2007, 102, 1. 123. W. D. Chemelewski, N. T. Hahn and C. B. Mullins, J.
Phys. Chem. C, 2012, 116, 5255−5261.
30 103. R. C. Chikate, K. Jun, C. V. Rode, Polyhedron, 2008, 27, 124. Y. Liang, C. S. Enache and R. Van de Krol, Int. J.
933. (a) D. Flak, A. Braun, K. A. Michalow, J. Wyrwa, 80 Photoenergy , 2008, 2008, 739864.
M. Parlinska-Wojtan, T. Graule and M. Rekas, J. 125. G. Wang, Y. Ling, D. A. Wheeler, K. E. N. George, K.
Nanosci. Nanotechno. 2012, 12, 6401. Horsley, C. Heske, J. Z. Zhang and Y. Li, Nano Lett.,
104. P. Deb, A. Basumallick, D. Sen, S. Mazumder, B. K. Nath 2011, 11, 3503.
35 and D .Das, Phil. Mag.Letters. 2006, 86, 491. 126. X. Lian, X. Yang, S. Liu, Y. Xu, C. Jiang, J. Chen and R.
105. W. Shenton, T. Douglas, M. Young, G. Stubbs and S. 85 Wang, Appl. Surf. Sci., 2012, 258, 2307.
Mann, Adv. Mater.1999, 11, 253. 127. J. A. Glasscock, P. R. F. Barnes, I. C. Plumb and N.
106. H. Cao, G. Wang, J. H. Warner and A. A. R. Watt, Savvides, J. Phys. Chem. C, 2007, 111, 16477.
Appl.Phys.Lett.2008, 92, 013110. 128. M. Zhang, W. Luo, Z. Li, T. Yu and Z. Zou, Appl. Phys.
Lett. 2010, 97, 042105.
40 107. J. Frydrych , L. Machala , M. Hermanek , I. Medrik , M. 90 129. S. Kumari , A. P. Singh, Sonal, D. Deva , R. Shrivastav ,
Mashlan, J. Tucek , J. Pechousek and V. K. Sharma, Thin S. Dass and V. R. Satsangi, Int. J. Hydrogen Energ.,
Solid Films, 2010, 518, 5916. 2012, 35, 3985.
130. J. He and B. A. Parkinson, ACS Comb. Sci. 2011, 13,
108. M. B. Sahana, C. Sudakar, G. Setzler, A. Dixit, J. S. 399.
Thakur, G. Lawes, R. Naik, V. M. Naik, and P. P. 95 131. P. Kumar , P. Sharma , R. Shrivastav , S. Dass and V. R.
45 Vaishnava, Appl. Phys. Lett., 2008, 93, 231909. Satsangi, Int. J. Hydrogen Energ.,2011, 36, 2777.
132. A. Kleiman-Shwarsctein, Y. Hu, A. J. Forman, G. D.
109. J. Cao, T. Kako, N. Kikugawa and J. Ye, J. Phys. D:
Stucky and E. W. McFarland, J. Phys. Chem. C,2008,
Appl. Phys. 2010, 43, 325101
112, 15900.
110. S. U. M. Khan and J. Akikusa, J. Phys. Chem. B, 1999, 100 133. A. Bak, W. Choi and H. Park, Appl. Catal B: Environ.,
103, 7184. 2011, 110, 207.
134. Y. Liu, Y. Yu and W. Zhang, Electrochim. Acta 2012,
59,121.
 135. Y. Hu, A. Kleiman-Shwarsctein, A. J. Forman, D. Hazen, 159. D. Gust, T. A. Moore and A. L. Moore, Faraday
J. Park and E. W. McFarland, Chem. Mater., 2008, 20, Discuss., 2012, 155, 9.
3803. 160. P.D. Tran, L. H. Wong, J. Barber and J.S.C Loo, Energy.
136. H. L. Sanchez, H. Steinfink and H. S. White, J. Solid 60 Environ. Sci., 2012, 5, 5902.
5 Sate Chem., 1982, 41, 90. 161. E. Reisner, J. C. Fontecilla-Camps and F. A. Armstrong,
137. T. Hisatomi, F. Le Formal, M. Cornuz, J. Brillet, N. Chem. Commun., 2009, 550.
Tetreault, K. Sivula and M. Gratzel, Energy Environ. 162. E. Reisner, D. J. Powell, C. Cavazza, J. C. Fontecilla-
Sci.,2011, 4, 2512. Camps and F. A. Armstrong, J. Am. Chem. Soc., 2009,
138. Y.Hu, A. Kleiman-Shwarsctein, G. D. Stucky and E. W. 65 131,18457.
10 McFarland, Chem. Comm., 2009, 2652. 163. L. M. Utschig, N. M. Dimitrijevic, O. G. Poluektov, S. D.
139. A. Kleiman-Shwarsctein, Y. Hu, G. D. Stucky and E. W. Chemerisov, K. L. Mulfort, and David M. Tiede, J. Phys.
McFarland, Electrochem. Comm., 2009, 11, 1150. Chem. Lett., 2011, 2, 236.
140. F. Le Formal, M. Gra¨tzel, and K. Sivula, Adv. Funct. 164. H. Tributsch and M. Calvin, Photochem. Photobiol.,
Mater. 2010, 20, 1099. 70 1971, 14, 95.
15 141. J. Brillet, M. Gratzel and K. Sivula, Nano Lett. 2010, 10, 165. E. Rabinowitch. J. Chem. Phys., 1940, 8, 560.
4155. 166. F. K. Fong and N. Winograd, J. Am. Chem. Soc.,1976, 98,
142. S. David Tilley, M. Cornuz, K. Sivula and M. Gratzel, 2287.
Angew. Chem. Int. Ed. 2010, 49, 1. 167. Y. Amao, N. Shuto, K. Furuno, A. Obata, Y. Fuchino, K.
143. H. Dotan, K. Sivula, M. Gratzel, A. Rothschild and S. C.75 Uemura, T. Kajino, T. Sekito, S. Iwai, Y. Miyamotod and
20 Warren, Energy Environ. Sci., 2011, 4, 958. M. Matsuda, Faraday Discuss., 2012, 155, 289.
144. A. Memar, W. Ramli, W. Daud , S. Hosseini , E. Eftekhari 168. H. Ochiai, H. Shibata, Y. Sawa and T. Katoh, Proc. Natl.
and L. J. Minggu, Sol. Energ.,2010, 84, 1538. Acad. Sci. USA, 1980, 77, 2442.
145. K. Sivula, F. Le Formal and M. Grätzel, Chem. 169. M. Kato, T. Cardona, A. W. Rutherford and E. Reisner, J.
Mater., 2009, 21, 2862. 80 Am. Chem. Soc., 2012, 134, 8332.
25 146. (a) D.A. Wheeler, G. Wang, Y. Ling, Y. Li and J.Z. 170. C. M. Eggleston, N. Khare and D. M. Lovelace, J.
Zhang, Energy & Environmental Science 2012, 5, 6682. Electron Spectrosc. 2006, 150, 220.
(b) L. Li, Y. Yu, F. Meng, Y. Tan, R. J. Hamers and S. 171. D. K. Bora, E. A. Rozhkova, K. Gazda - Schrantz, P. P.
Jin, Nano Lett., 2012, 12, 724−731 Wyss, A. Braun, T. Graule and E. C. Constable, Adv.
147. Y. Ling, G.Wang, D. A. Wheeler, J. Z. Zhang and Y. Li, 85 Funct. Mater., 2012, 22, 490.
30 Nano Lett., 2011, 11, 2119. 172. N. Khare, C. M. Eggleston, D. M. Lovelace and S. W.
148. D. Qin, C. Tao, S. In, Z,Yang, T. E. Mallouk, N. Bao and Boese, J. Colloid Interf. Sci., 2006, 303, 404.
C. A. Grimes, Energ. Fuels, 2011, 25, 5257. 173. K. H. Kung and M. B. Mcbride, Clays Clay Miner., 1988,
149. (a) Y. Lin, S. Zhou, S. W. Sheehan and D. Wang, J. Am. 36, 303.
Chem. Soc., 2011, 133, 2398. (b) Y. Lin, G. Yuan, S. 90 174. D. S. Berns and R. McColl, Chem. Rev.1989, 89, 807
35 Sheehan, S. Zhou and D. Wang, Energy & Environmental 175. S. Zhang, J. Xie, J. Zhang, J. Zhao and L. Jiang, Biochim.
Science, 2011, 4, 4862. Biophys. Acta,1999, 205, 1426.
150. T. Vincent, M. Gross, H. Dotan and A. Rothschild, Inter. 176. K. Kalyanasundaram and M. Gratzel, Coordin. Chem.
J. Hydrogen Energ., 2012, 37, 8102. Rev., 1998, 77, 347.
151. S. K. Mohapatra, S. E. John, S. Banerjee and M. Misra, 95 177. F. V. Herrera, P. Grez, R. Schrebler, L. A. Ballesteros, E.
40 Chem. Mater. 2009, 21, 3048. Muñoz, R. Córdova, H. Altamirano and E. A. Dalchiele,
152. D. K. Bora, A. Braun, R. Erni, G. Fortunato, T. Graule J. Electrochem. Soc., 2010, 157, D302.
and E. C. Constable, Chem. Mater. 2011, 23, 2051. 178. 178. Z. Huang, Y. Lin, X. Xiang, W. Rodriguez-Cordoba,
153. D. K. Bora and P. Deb, Nanoscale Res. Lett., 2009, K.J. McDonald, K.S. Hagen, K.-S. Choi, B. S.
4,138. 100 Brunschwig, D.G. Musaev, C.L. Hill, D. Wang and T.
45 154. M. Das, D. Mishra, T. K. Maiti, A. Basak and P Lian, Energy & Environmental Science, 2012, 5, 8923.
Pramanik, Nanotechnology, 2008, 19, 415101. 179. 179. K. Gajda-Schrantz, S. Tymen, F. Boudoire, R. Toth,
155. D. Liu, W. Wu, X. Chen, S. Wen, X. Zhang, Q. Ding, D.K. Bora, W. Calvet, M. Grätzel, E. C. Constable, A.
G. Teng and N. Gu, Nanoscale, 2012, 4, 2306. Braun, Phys. Chem. Chem. Phys. in press
156. R. Narain, M. Gonzales, A. S. Hoffman, P. S. Stayton and 105 http://pubs.rsc.org/en/content/articlelanding/2012/cp/C2CP425
50 K. M. Krishnan, Langmuir, 2007, 23, 6299. 97A
157. L. Liu, H. Dietsch, P. Schurtenberger and M. Yan,
Bioconjugate Chem., 2009, 20, 1349.
158. A. Z. Wang, V. Bagalkot, C. C. Vasilliou, F. Gu, F.
Alexis, L. Zhang, M. Shaikh, K. Yuet, M. J. Cima, R.
55 Langer, P. W. Kantoff, N. H. Bander, S. Jon and O.C.
Farokhzad, ChemMedChem, 2008, 3, 1311.