of 10
A taxonomy for solar fuels generators
Adam C. Nielander,
Matthew R. Shaner,
Kimberly M. Papadantonakis,
Sonja A. Francis and Nathan S. Lewis
*
A number of approaches to solar fuels generation are being developed, each of which has associated
advantages and challenges. Many of these solar fuels generators are identi
fi
ed as
photoelectrochemical
cells
even though these systems collectively operate based on a suite of fundamentally di
ff
erent
physical principles. To facilitate appropriate comparisons between solar fuels generators, as well as to
enable concise and consistent identi
fi
cation of the state-of-the-art for designs based on comparable
operating principles, we have developed a taxonomy and nomenclature for solar fuels generators based
on the source of the asymmetry that separates photogenerated electrons and holes. Three basic device
types have been identi
fi
ed: photovoltaic cells, photoelectrochemical cells, and particulate/molecular
photocatalysts. We outline the advantages and technological challenges associated with each type, and
provide illustrative examples for each approach as well as for hybrid approaches.
Broader context
Solar fuels generators are devices that harness energy from sunlight to drive the synthesis of chemical fuels. A number of approaches to solar fuels ge
neration are
being pursued, many of which can be di
ff
erentiated by the physical principles on which they are based. Herein, we propose a nomenclature and taxonomy based
on three basic device types: photovoltaic cells, photoelectrochemical cells, and photoelectrosynthetic particulate/molecular photocatalysts
. An understanding of
the inherent operating principles and the advantages and challenges associated with each of these device types will facilitate clear comparisons be
tween devices
as well as help guide research e
ff
orts toward improving these devices and achieving the ultimate goal of sustainable fuel production.
Introduction
The development of an arti
cial photosynthetic process,
whereby the energy from sunlight is captured and stored in the
chemical bonds of a fuel, has been an active area of research for
decades. This
eld of research, however, has recently under-
gone rapid expansion due to the promise of a scalable solar
fuels generator that would provide a carbon-neutral source of
energy capable of addressing concerns about the impact of
carbon emissions on climate while providing a measure of
environmental and energy security. Researchers have developed
a diverse set of designs for solar fuels generators (Fig. 1), each of
which presents unique challenges associated with the research
and development required to obtain a fully operational system.
Furthermore, the maturity of the technologies being imple-
mented in the various designs varies widely. Despite these
di
ff
erences, a variety of solar fuels generators are o
en grouped
together and denoted as
photoelectrochemical cells
. The
focus of this Opinion is to establish a di
ff
erentiating
Fig. 1
Illustrations of the di
ff
erent categories of solar fuels generators.
(a) Semiconductor/electrolyte junction in the dark and prior to equil-
ibration in which the photovoltage and photocurrent are determined
in whole or in part by the di
ff
erence between Fermi level of the
semiconductor (
E
F
) and the electrochemical potential of the electro-
lyte solution (
E
redox
), denoted as
D
E
. (b) Semiconductor buried junction
in the dark and prior to equilibration in which the photovoltage and
photocurrent are determined by the di
ff
erence between the Fermi
levels (
E
F
) of the two solid-state contacting phases (
D
E
), shown here as
two semiconductors. The
D
E
is independent of any di
ff
erence
between the Fermi level of the solid contacting the electrolyte and the
electrochemical potential of the electrolyte. The highly doped phase
(in red) allows for ohmic contact between it and the contacting
electrolyte phase. (c) Particulate/molecular photocatalyts suspended
or dissolved in solution. Each unit individually absorbs light, generates
excited carriers and e
ff
ects the desired chemical reactions at the
particulate/molecular electrolyte interface.
Beckman Institute, Kavli Nanoscience Institute, Joint Center for Arti
cial
Photosynthesis, 210 Noyes Laboratory, 127-72 Division of Chemistry and Chemical
Engineering, California Institute of Technology, Pasadena, CA 91125, USA. E-mail:
nslewis@caltech.edu
These authors contributed equally to this work.
Cite this:
Energy Environ. Sci.
,2015,
8
,
16
Received 18th July 2014
Accepted 19th September 2014
DOI: 10.1039/c4ee02251c
www.rsc.org/ees
16
|
Energy Environ. Sci.
,2015,
8
,16
25
This journal is © The Royal Society of Chemistry 2015
Energy &
Environmental
Science
OPINION
Published on 19 September 2014. Downloaded on 15/01/2015 15:34:38.
View Article Online
View Journal
| View Issue
nomenclature and taxonomy for solar fuels generators that
clearly identi
es the principles underlying the designs. We
hope that adoption of this taxonomy (Scheme 1) will bring
clarity and precision to discussions and comparisons of solar
fuels devices while facilitating concise and consistent identi
-
cation of the research challenges and state-of-the-art for each
type of system.
All solar fuels generators require an electrical asymmetry to
separate and transport photogenerated charge carriers vecto-
rially.
1
4
Without vectorial separation and transport, the charge
carriers, and thus the chemical products, would have no net
directionality and thus would undergo no net separation.
Hence, deleterious recombination of charge carriers and/or a
loss of chemical potential in the resulting fuel/oxidant mixture
would result. The required vectorial separation can be e
ff
ected
by chemical and/or electrical potential gradients as well as by
kinetic asymmetries at the interface between two unlike mate-
rials.
1
4
We refer to this interface as a
junction
. We note that
our usage of the term
junction
di
ff
erentiates such an interface
from an interface between two unlike materials that does not
result in an asymmetry which produces a vectorial charge
separation.
5
We propose that the various solar-fuels generators
Scheme 1
A taxonomy for the classi
fi
cation of solar energy converters. For a device of interest, identify
n
(total number of junctions),
m
(number
of semiconductor/electrolyte junctions), and
l
(number of buried junctions), then proceed through the
fl
ow chart to determine the appropriate
name for the device. Note that the taxonomy does not address devices designed to use light to drive exergonic processes.
This journal is © The Royal Society of Chemistry 2015
Energy Environ. Sci.
,2015,
8
,16
25 |
17
Opinion
Energy & Environmental Science
Published on 19 September 2014. Downloaded on 15/01/2015 15:34:38.
View Article Online
can be di
ff
erentiated at a fundamental level based on the
underlying principles used to accomplish vectorial charge
separation and by the method in which the separated charge is
used to e
ff
ect the synthesis of chemical fuels.
Photovoltaic cells
One fundamentally identi
able approach to charge separation
in solar fuels devices is through the use of solid-state, or buried,
junctions. Buried junctions are exclusively formed at the inter-
face between two electronic conductors (as opposed to ionic
conductors,
vide infra
) and are the basis for the operation of
photovoltaic
(PV)
cells
(Fig. 2a).
6
8
In a device utilizing a buried
junction, the photovoltage and photocurrent produced in the
presence of illumination arise from charge separation mediated
by a di
ff
erence in electrochemical potential (Fig. 1b) and/or by a
di
ff
erence in charge-transfer kinetics between two unlike solids
that are in mutual electrical contact. The photocurrent
vs.
voltage behavior of a PV cell is independent of the character of
any solid/electrolyte interfaces in the system. Hence, measure-
ments of the photocurrent
voltage characteristics of the PV cell
can be performed independently of any electrochemical reac-
tion, and can be used in concert with the current
voltage
characteristics of various electrocatalysts to accurately predict
the performance of a complete solar fuels generator that is
based on a PV cell. PV cells will also produce the identical
photocurrent
voltage behavior when both terminals of the
device are contacted with wires connected to electrocatalysts
vs.
when all of the components of the structure (light absorbers
and electrocatalysts) are integrated, contacted intimately, and
immersed in an electrolyte solution. The operating principles of
photovoltaic electrodes have been well documented for incor-
poration into full PV cells that either produce electricity or
fuels.
3
PV cells that produce electricity are referred to as
solar elec-
tric cells
and are widely available commercially. PV cells that
produce fuels are referred to as
PV-biased electrosynthetic cells
and can consist of any number of buried junctions arranged
electrically in series with electrocatalysts submerged in an
electrolyte. The electrocatalysts may or may not be in physical
contact with the PV electrodes, but in all such systems the
photovoltage generated by the structure is independent of the
nature of the electrocatalyst/electrolyte interface. Examples of
PV-biased electrosynthetic cells include AlGaAs/Si tandem
structures,
9
amorphous hydrogenated Si (a-Si:H) triple-junction
structures,
10
12
triple-junction structures based on CuInGaSe
2
(Fig. 2b
d),
13
and n-Si/SiO
x
/In-doped tin oxide (ITO) structures.
14
The advantages of PV-based solar fuels generators are the
high reported solar-to-fuels e
ffi
ciencies and the independence
of the power-producing junction with respect to the formal
potential for the reactions of interest.
15
The challenges associ-
ated with PV-based cells include achieving a cost advantage for a
system with the functioning photovoltaic cell immersed in the
electrolyte, relative to a system that utilizes a discrete photo-
voltaic cell in dry conditions wired to a discrete fuel-forming
device, as well as
nding catalyst/electrolyte interfaces that are
transparent, conductive, and stable under operational, fuel-
forming conditions.
11,12,15
18
Thus, the key research needs
involve the development of cost-competitive photovoltaic cells,
the integration of components, discovery of materials, devel-
opment of low-cost fabrication methods, and the stabilization
of electrodes through the use of materials that act as trans-
parent and conductive protecting layers.
Photoelectrochemical cells
Another fundamentally identi
able approach to e
ff
ect the sepa-
ration of charge carriers is through the use of a solid/ionic-
conductor junction. Devices utilizing solid/ionic-conductor
junctions, also referred to as solid/electrolyte junctions, are called
photoelectrochemical
(PEC)
cells
(Fig. 3a). The solid in a PEC cell is
commonly a semiconductor and may or may not have an
attached photosensitizer. Other solids, including metals such as
platinum and mercury, have also been observed to produce a
photovoltage at a solid/electrolyte interface when the appropriate
Fig. 2
Illustrations of photovoltaic cells in the dark after equilibration,
with the physical position of the buried junction (BJ), the space charge
layer (SCL), and the electric potentials of the conduction and valence
bands shown in each diagram. The expected direction of electron
fl
ow
under illumination is also indicated. (a) A general photovoltaic elec-
trode in a complete photovoltaic cell with generic half reactions at
each electrode. (b) An example of a PV-biased electrosynthetic cell
that uses a monolithic tandem junction cell made of AlGaAs/Si
coupled to Pt and RuO
2
electrocatalysts (not shown) to split water. (c)
An example of a PV-biased electrosynthetic cell that uses a monolithic
triple junction cell made of amorphous hydrogenated Si and alloys
with Ge, which has been coupled to hydrogen- and oxygen-evolution
catalysts (not shown) to split water. (d) A recent example of a PV-
biased electrosynthetic cell that uses three separate side-by-side
CuInGaSe
2
single-junction cells coupled to electrocatalysts (not
shown) to split water.
18
|
Energy Environ. Sci.
,2015,
8
,16
25
This journal is © The Royal Society of Chemistry 2015
Energy & Environmental Science
Opinion
Published on 19 September 2014. Downloaded on 15/01/2015 15:34:38.
View Article Online
electrolyte is present.
19,20
In a device utilizing a solid/electrolyte
junction, the photovoltage and photocurrent produced in the
presence of light arise from di
ff
erences in the electrochemical
potentials of the solid and the electrolyte as well as from asym-
metries in the charge-transfer kinetics for electrons and holes
across the junction. The operating principles of photoelectrodes
have been well documented for incorporation into full PEC cells
that either produce electricity or fuels.
21
The properties of such
photoelectrodes are determined routinely from a conventional
three-electrode potentiostatic experiment using a half-cell
con
guration, with the understanding that the photoelectrode
can be incorporated into an operational, two-electrode, full PEC
cell. Unlike PV cells, for a given PEC-based solar fuels generator,
photocurrent
voltage measurements cannot be made indepen-
dently of the reaction of interest.
PEC cells that utilize a semiconducting electrode can consist
of one photoelectrode that has a semiconductor/electrolyte
junction, in conjunction with a
dark
counter electrode
(Fig. 3a); of two photoelectrodes, each with a semiconductor/
electrolyte junction (Fig. 3b); or of a monolithically integrated
combination of two photoelectrodes in a single structure that
performs both the anodic and cathodic half-reactions simulta-
neously (Fig. 3c).
PEC cells that only produce electricity are referred to as
regenerative photoelectrochemical cells
(Fig. 4a), because the
species that is reduced or oxidized at the working photo-
electrode is regenerated at the counter electrode, ideally
yielding zero net change in the composition of the
solution.
4,7,8,22
PEC cells that produce fuels at the semi-
conductor/electrolyte junction are referred to as
photo-
electrosynthetic cells
(Fig. 4b).
7,23,24
An example of a regenerative
PEC cell is the n-Si/CH
3
OH-1,1
0
dimethylferrocene
+/0
/ITO cell
(Fig. 4c, n-Si is the photoelectrode).
25
Dye-sensitized solar cells
(DSSCs) are also commonly operated as regenerative PEC cells.
26
Examples of photoelectrosynthetic PEC cells include n-SrTiO
3
/
NaOH(aq)/Pt cells for water splitting (Fig. 4d, SrTiO
3
is the
photoelectrode),
27
n-MoS
2
cells for the production of H
2
and I
2
from HI(aq) (Fig. 4e, n-MoS
2
is the photoelectrode),
28
and
DSSC's including TiO
2
photosensitized with a catalytic molec-
ular [((PO
3
H
2
)
2
bpy)
2
Ru
a
(4-Mebpy-4-bimpy)Ru
b
(tpy)(OH
2
)]
4+
unit
for water splitting, as well as others.
29
31
The product of coupling a regenerative PEC cell to metallic
electrodes produces a
PEC-biased electrosynthetic cell
, whereas
the product of coupling a regenerative PEC cell to a photo-
electrosynthetic PEC cell is referred to as a
PEC-biased photo-
electrosynthetic cell
. Photoelectrochemical cells, like
photovoltaic cells, can be used to bias both PEC and PV cells to
assist in fuel formation. An example of a PEC-biased photo-
electrosynthetic cell is a DSSC placed electrically in series with
an Fe
2
O
3
/electrolyte junction cell for water splitting (Fig. 4f).
32
Here, the DSSC is a free-standing, two-terminal device whose
photocurrent and photovoltage are independent of the fuel-
forming reactions of interest, but which operates as a PEC
nonetheless because the photocurrent and photovoltage are not
independent of the solution at the interface of the two terminals
of the DSSC itself.
The performance of photoelectrodes consisting of semi-
conductor/electrolyte junctions, in the absence of bulk semi-
conductor limitations, is determined by the energetics and
kinetics of the semiconductor/electrolyte interface. The inter-
facial energetics determine the photovoltage through the
di
ff
erence between the formal potential of the fuel-forming
reaction of interest and the electrochemical potential of the
semiconductor,
18,33,34
and also determine the driving force
needed to produce a given current density. Commonly, an
electrocatalyst is incorporated at the semiconductor/electrolyte
interface to improve the interfacial charge-transfer kinetics;
however, for the device to remain categorized as a PEC cell, the
nature of the electrolyte must a
ff
ect the performance of the
cell.
35
Examples of PEC cells with electrocatalysts incorporated
at the semiconductor/electrolyte interface include H
2
-evolving
photocathodes made from metal islands or thin metallic
lms
on p-Si or p-InP photoelectrodes, because the work function of
the metal, and thus the barrier height at the semiconductor
surface, depends on the concentration of H
2
in the electro-
lyte.
36,37
Semiconductor/electrolyte junctions with ion-perme-
able, redox-active electrocatalysts would also be considered PEC
cells because of the electrolyte-dependent behavior of the
device.
35
In addition, recent progress on stabilization schemes
based on thin coatings on the surface of the semiconductor has
produced examples of photoelectrodes in which the solution
potential a
ff
ects the photovoltage even though the photo-
electrode is not in direct physical contact with the solution.
38,39
Conversely, electrocatalysts deposited on semiconductors, such
as CoPi on Fe
2
O
3
, are reported to convert what would otherwise
Fig. 3
Illustrations of photoelectrochemical cells that e
ff
ect two
general half-reactions with (a) one photoelectrode and a dark elec-
trode, (b) two separate (dual) photoelectrodes, and (c) a monolithic
structure with two junctions oriented in series with respect to the
incoming light (tandem photoelectrodes). The physical location of the
semiconductor/electrolyte junction (SEJ), the space charge layer
(SCL), and the electric potentials of the conduction and valence bands
are shown in each diagram.
This journal is © The Royal Society of Chemistry 2015
Energy Environ. Sci.
,2015,
8
,16
25 |
19
Opinion
Energy & Environmental Science
Published on 19 September 2014. Downloaded on 15/01/2015 15:34:38.
View Article Online
be photoelectrosynthetic cells into photovoltaic electrosynthetic
cells, by formation of a Schottky junction at the semiconductor/
catalyst junction.
40
42
Careful evaluation is o
en necessary to
determine whether a device is a PV or PEC cell when electro-
catalysts are present on the surface. Data including the current
voltage characteristics of the catalyst alone, the photocurrent
Fig. 4
Illustrations of photoelectrochemical cells in the dark after equilibration, with the physical position of the semiconductor/electrolyte
junction (SEJ), the space charge layer (SCL), and the electric potentials of the conduction and valence bands shown in each diagram. The
expected direction of electron
fl
ow under illumination is indicated. The axes in each panel are the same as those in panel a. (a) A general
regenerative photoelectrochemical cell with one photoelectrode and one dark electrode e
ff
ecting two half-reactions that collectively yield zero
net reaction. This cell produces electrical power, as indicated by the load in the circuit. (b) A general photoelectrosynthetic cell that splits wate
ras
an example of solar-driven catalysis of a chemical reaction with
D
G
> 0. (c) An example of a regenerative photoelectrochemical cell that uses an
n-Si photoelectrode and an ITO dark electrode to produce electrical power using a CH
3
OH-1,1
dimethylferrocene
+/0
redox couple. (d) An
example of a photoelectrosynthetic cell that uses an n-SrTiO
3
photoelectrode and a Pt dark electrode to split water. (e) An example of a
photoelectrosynthetic cell that uses an n-MoS
2
photoelectrode and a Pt dark electrode to e
ff
ect the unassisted splitting of HI. (f) An example of a
PEC-biased photoelectrosynthetic cell that uses a monolithic DSSC/n-Fe
2
O
3
combination to split water.
20
|
Energy Environ. Sci.
,2015,
8
,16
25
This journal is © The Royal Society of Chemistry 2015
Energy & Environmental Science
Opinion
Published on 19 September 2014. Downloaded on 15/01/2015 15:34:38.
View Article Online
voltage characteristics of the semiconductor with and without
the presence of the electrocatalyst, the photocurrent
voltage
behavior of the semiconductor with and without electrocatalyst
in contact with electrolytes of varying composition and elec-
trochemical potential, and laser spectroscopic data on the
electron
hole recombination mechanism in the presence or
absence of electrocatalyst may be necessary to ascertain
whether such a system is properly classi
ed as a PV or PEC cell.
The principal advantages of PEC cells are their simplicity of
fabrication and the
nding that inexpensive polycrystalline
semiconductor/electrolyte junctions can o
en perform nearly
as well as their single crystalline counterparts.
43
45
The chal-
lenges associated with PEC cells include obtaining a combina-
tion of materials that are operationally stable and also possess
appropriate interfacial energetics and band gaps, as well as the
development and integration of electrocatalysts into the semi-
conductor/electrolyte junction. Thus, the key research needs for
solar fuels generators based on PEC cells involve the discovery
and development of semiconducting materials that possess
both the proper band gaps for e
ff
ective sunlight absorption and
well-positioned band energetics, and the development of
methods for incorporating e
ffi
cient electrocatalysts into semi-
conductor/electrolyte interfaces that are stable under opera-
tional, fuel-forming conditions.
18,46
49
Photovoltaic-biased
photoelectrochemical cells
The product of coupling a PV cell with a PEC cell, resulting in a
cell that contains both a buried junction and a semiconductor/
electrolyte junction, is a
PV-biased PEC cell
(Fig. 5a). In this
hybrid approach, the advantages of both cells are combined
through increased
exibility in materials availability. Like their
parent cells, PV-biased PEC cells can produce electricity or fuel.
PV-biased PEC cells that produce electricity are referred to as
a
PV-biased regenerative PEC cell
. PV-biased PEC cells that
produce fuels and that include at least one buried junction may
fall into a number of categories, which are systematically named
based on whether fuel formation occurs at a solid/electrolyte
junction in the device and the presence or absence of additional
two-terminal regenerative PEC cells.
PV-biased photo-
electrosynthetic cells
are PV-biased PEC cells in which fuel
formation occurs at the solid/electrolyte junction. PV-biased
PEC cells that produce fuels that are formed away from a solid/
electrolyte junction, but include at least one isolated regenera-
tive PEC cell, are referred to as
Regenerative PEC- and PV-biased
electrosynthetic cells
. PV-biased PEC cells that include at least
one isolated regenerative PEC cell, but that produce fuels that
are formed at a solid/electrolyte junction, are referred to as
Regenerative PEC- and PV-biased photoelectrosynthetic cells.
Examples of PV-biased PEC cells include the
Turner Cell
,a
GaAs buried junction electrically in series and monolithically
integrated with a p-GaInP
2
/electrolyte junction (Fig. 5b), as well
as an a-Si:H PV cell electrically in series with a BiVO
4
/electrolyte
junction (Fig. 5c) and the PEC cells o
en referred to as septum-
based PEC cells.
50
54
Photoelectrosynthetic particulate/
molecular photocatalysts
Both the buried junction and the semiconductor/electrolyte
junction motifs can be employed when the semiconducting
material is employed in a dispersed particulate form as opposed
to a solid electrode (Fig. 6). In these particulate systems, the
photovoltage and photocurrent that drive the interfacial elec-
trochemical reactions in the presence of illumination are
developed as a result of semiconductor/electrolyte and/or
buried junctions in a single discrete particle unit that generally
contains separate co-catalysts for each half-reaction.
55
Although
in theory one could distinguish between particles utilizing
buried and semiconductor/electrolyte junctions in the same
way as for the PEC and PV cells, in practice, these two types of
systems are di
ffi
cult to distinguish experimentally. A compar-
ison of the photovoltage produced by a particle in solution with
that measured across a particle removed from solution may be
di
ffi
cult or impossible to perform, due to the small size of the
particles and the resulting e
ff
ective absence of addressable
Fig. 5
Illustrations of photovoltaic-biased photoelectrochemical cells
in the dark after equilibration, with the physical position of the semi-
conductor/electrolyte junction (SEJ), the space charge layer (SCL), the
buried junction (BJ), and the electric potentials of the conduction and
valence bands shown in each diagram. The expected direction of
electron
fl
ow under illumination is indicated. The axes in each panel
are the same as those in panel a. (a) A general photovoltaic-biased
photoelectrochemical cell comprising one photovoltaic electrode and
one photoelectrochemical electrode that e
ff
ects two general half-
reactions. (b) An example of a photovoltaic-biased photo-
electrochemical cell that uses tandem photoelectrodes, one utilizing a
GaInP/electrolyte junction and the other utilizing a GaAs buried
junction, to drive water splitting. (c) A recent example of a photovol-
taic-biased photoelectrochemical cell that uses tandem photo-
electrodes, one utilizing a BiVO
4
/electrolyte junction and the other
utilizing an amorphous hydrogenated Si buried junction, to drive water
splitting.
This journal is © The Royal Society of Chemistry 2015
Energy Environ. Sci.
,2015,
8
,16
25 |
21
Opinion
Energy & Environmental Science
Published on 19 September 2014. Downloaded on 15/01/2015 15:34:38.
View Article Online
electrodes. Indirect measurements of the photocurrent and/or
photovoltage under varying conditions may be obtained by
correlating changes in the amount of products formed by the
light-driven reaction with various solution compositions, but
accurate measurements of the products will be hindered by
product crossover and incompatible catalysts. The particulate
versions of PV and PEC cells, as well as the related photo-driven
molecular photocatalysts wherein inorganic molecular
compounds are dispersed in solution, share many of the same
research challenges as their parent categories, with the added
challenge of developing methods to physically separate the
products of the fuel-forming reactions. The term cell does not
apply to particulate schemes that employ neither addressable
electrodes nor a built-in means to enforce the separation of
products. For these reasons, we consider all three of these
strategies to comprise members of the general category of
photoelectrosynthetic particulate/molecular photocatalysts
.
7,56,57
An example of photoelectrosynthetic particulate photo-
catalysts are CdS particles in contact with TiO
2
particles, with an
electrical asymmetry at the CdS/TiO
2
interface.
57
59
Other
examples include a NiO
SrTiO
3
photocatalyst capable of
concomitantly evolving H
2
and O
2
, as well as a number of metal
nitrides, oxides, and oxynitrides (
e.g.
ZrO
2
, TaON, Ta
3
N
5
,
WO
3
).
57
Similarly, the performance of a photoelectrosynthetic
molecular photocatalyst is based either on monomolecular
photochemical processes or on coupled photoelectrochemical
photochemical or photochemical-dark reactions in an indi-
vidual molecular unit. Examples of photoelectrosynthetic
molecular cells include light-driven water splitting by UV irra-
diation of aqueous Ce(
III
)/Ce(
IV
) solutions;
60
the use of molecular
triads or tetrads coupled to nanoparticulate or molecular elec-
trocatalysts for fuel production;
61
the coupling of molecular
catalysts to photoactive proteins;
62,63
and related systems.
64,65
The principal advantages of particulate/molecular photo-
catalysts are the simplicity of the photocatalysts relative to other
approaches and the associated low predicted system cost, with a
recent technoeconomic analysis suggesting that systems based
on particulate/molecular photocatalysts could be signi
cantly
less expensive than electrode-based systems when deployed at
scale.
15
The challenges facing development of systems from
photoelectrosynthetic particulate/molecular photocatalysts
involve stabilizing all of the components; addressing safety
concerns arising from the production of explosive mixtures of
stoichiometric fuel products; and controlling undesired
recombination processes to realize high steady-state quantum
yields for net fuel production. Speci
c undesired processes
include photogenerated electrons reducing key surface-bound
intermediates, intermediates in solution, or products of the
oxidation of water to O
2
, as well as photogenerated holes
participating in analogous oxidation reactions, and the spon-
taneous recombination of the fuels facilitated by contact with
the co-catalysts at any location in the system.
Discussion
Both PV and PEC cells can be structured with multiple junctions
to optimize the theoretical maximum e
ffi
ciency for a given fuel-
forming reaction.
66
Single-junction cells are best suited for fuel-
forming reactions that require operating voltages near or below
the

1 V maximum power point of the single-junction devices
that have the highest reported energy-conversion e
ffi
ciency.
67
Fuel-forming reactions that require larger voltages also require
more junctions to better utilize the solar spectrum, with the
optimal number of junctions being dependent on the voltage
required for the operating current density. Hence, the
maximum realizable e
ffi
ciencies for water splitting are generally
obtained with a tandem-absorber cell, where two light
absorbers with appropriately tuned band gaps are arranged in
series with respect to the incident light.
9,53
Additional junctions
can increase the e
ffi
ciency of solar devices when the semi-
conductors have carefully selected band gaps.
68
Triple-junction
cells utilizing a single semiconductor or two semiconductors
have also been used to e
ff
ect solar-driven water splitting when
related double junction devices were unable to generate su
ffi
-
cient voltage.
11,12,51
When the same semiconductor is used to
form multiple junctions, the cells su
ff
er from a loss of current to
achieve the necessary voltage for water splitting.
Advanced structuring of PV- and PEC-based solar fuels
generators can o
ff
er additional e
ffi
ciency gains for systems,
including those for which all of the components are in contact
with the electrolyte. One example of advanced structuring is an
array of p-Si microwires that have radial n
+
-doped emitter
regions, with an electrocatalyst placed in speci
c physical
locations between or along the surfaces of the microwires
(Fig. 7).
69
Some ambiguity exists regarding the classi
cation of
such a system. The mechanism of charge separation is a buried
junction, and thus the device falls into the category of photo-
voltaic cells. However, although in concept a conformal elec-
trical contact could be made to the microwires, with the
resulting electrical current then passed to another identically
microstructured conductive electrode that possessed the spatial
distribution, loading, and resulting activity of the
Fig. 6
Illustration of a photocatalytic particle in the dark after equili-
bration, with the physical position of the semiconductor/electrolyte
junction (SEJ) and the electric potentials of the conduction and
valence bands shown in the diagram. This photocatalytic particle is
shown performing water splitting with two separate co-catalysts
attached to its surface. Instead of a semiconductor/electrolyte junc-
tion the particle could utilize a buried junction. In practice, multiple
particles are suspended in an electrolyte.
22
|
Energy Environ. Sci.
,2015,
8
,16
25
This journal is © The Royal Society of Chemistry 2015
Energy & Environmental Science
Opinion
Published on 19 September 2014. Downloaded on 15/01/2015 15:34:38.
View Article Online
electrocatalyst in the integrated structure, separation of the
integrated system into essentially identically functioning
discrete components would be di
ffi
cult to accomplish in prac-
tice. Because the performance of the device critically depends
on the details of, and the presence of, the absorber/electrolyte
junction, which acts in this case in a synergistic fashion with
respect to one or more other components of the integrated
system, designation of the device as a photoelectrochemical cell
might seem reasonable. Furthermore, if the microwires are
removed from the substrate and embedded in an immobilizing
membrane, they may be deemed similar to a photo-
electrosynthetic particulate photocatalyst. In this case, however,
the uniform particle orientation and built-in barrier for product
separation, would produce a photoelectrosynthetic particulate
cell rather than a photoelectrosynthetic particulate photo-
catalyst. This discussion serves to emphasize that while some
devices easily fall into a single taxonomic classi
cation and
therefore are subject to the research challenges associated with
that classi
cation, other devices may have characteristics of
multiple classi
cations with some or all of the related chal-
lenges, advantages or disadvantages.
Conclusions
Although researchers have developed diverse designs for solar
fuels generators based on a diverse set of underlying principles,
solar fuels generators are o
en grouped together and denoted
as
photoelectrochemical cells
. The purpose of this Opinion is
not to favor, or establish a bias or preference towards, any
speci
c design or approach. The di
ff
erent performance/cost/
function trade-o
ff
s associated with each approach ultimately
will determine which of these distinct technological approaches
to the development of solar fuels generators will prove viable in
the marketplace. Instead, we have described a taxonomy for
solar fuels generators that is based on the operating principles
underlying the designs, to bring clarity and precision to
discussions of research in the
eld of arti
cial photosynthesis
while facilitating concise and consistent identi
cation of the
research challenges and state-of-the-art for each type of system.
Acknowledgements
ACN acknowledges the NSF, Grant CHE-1214152, and the
National Science Foundation Graduate Research Fellowship for
support. KMP acknowledges support from the DOE, Grant DE-
FG02-03ER15483. MRS acknowledges the Resnick Sustainability
Institute for a graduate fellowship. This material is based upon
work performed by the Joint Center for Arti
cial Photosyn-
thesis, a DOE Energy Innovation Hub, supported through the
O
ffi
ce of Science of the U.S. Department of Energy under Award
Number DE-SC0004993.
References
1 A. J. Bard and M. A. Fox,
Acc. Chem. Res.
, 1995,
28
, 141
145.
2 S. J. Fonash,
Solar Cell Device Physics
, Elsevier Inc.,
Amsterdam, 2nd edn., 2010.
3 N. Serpone and E. Pelizzetti,
Photocatalysis: Fundamentals
and Applications
, John Wiley & Sons, Inc., Hoboken, NJ, 1989.
4 M. X. Tan, P. E. Laibinis, S. T. Nguyen, J. M. Kesselman,
C. E. Stanton and N. S. Lewis, in
Prog. Inorg. Chem.
, ed. K.
D. Karlin, John Wiley & Sons, Inc., 1994.
5 Within this Opinion, we speci
cally denote as
junctions
those interfaces whose electrical asymmetry leads to the
development of a photovoltage or photocurrent upon
illumination. In other contexts, the interface between two
materials that do not generate a photocurrent or
photovoltage under illumination, such as ohmic contacts,
may also be referred to as a
junction
.
6 The term
photovoltaic
has been used in a number of
contexts, including to denote any device that generates a
measurable photocurrent or photovoltage upon
illumination, including speci
cally photoelectrochemical
cells (ref. 7 and 8). This usage is defendable because both
PV and PEC cells exhibit a photovolatic e
ff
ect as
understood by the strict de
nition of the term
photovoltaic
. However, as alluded to in ref. 8,
photoelectrochemical cells are commonly di
ff
erentiated
from cells that utilize a junction between two electronic
conductors, to emphasis the di
ff
erent underlying
principles in the two di
ff
erent types of systems. The term
photovoltaic
is therefore used herein to refer to only those
devices that include a junction between electronic
conductors.
7 A. J. Bard, R. Memming and B. Miller,
Pure Appl. Chem.
, 1991,
63
, 569
596.
8
Semiconductor Electrochemistry and Photoelectrochemistry
, ed.
S. Licht, Wiley-VCH, Weinheim, Germany, 2002.
9 S. Licht, B. Wang, S. Mukerji, T. Soga, M. Umeno and
H. Tributsch,
J. Phys. Chem. B
, 2000,
104
, 8920
8924.
10 W. Ayers,
US Pat.
, 4,466,869, 1984.
11 S. Y. Reece, J. A. Hamel, K. Sung, T. D. Jarvi, A. J. Esswein,
J. J. H. Pijpers and D. G. Nocera,
Science
, 2011,
334
, 645
648.
12 R. E. Rocheleau, E. L. Miller and A. Misra,
Energy Fuels
, 1998,
12
,3
10.
13 T. J. Jacobsson, V. Fjallstrom, M. Sahlberg, M. Edo
ff
and
T. Edvinsson,
Energy Environ. Sci.
, 2013,
6
, 3676
3683.
Fig. 7
A next-generation photocathode that employs advanced
structuring to improve performance. This photocathode consists of an
array of Si microwires, each microwire with a buried junction and
connected to hydrogen-evolution catalysts (Ni
Mo). TiO
2
scattering
particles are located at the base of the microwire array to maintain high
catalytic activity while also maintaining high light absorption in the
semiconductor. This photocathode could be combined with a (photo)
anode to form a complete cell.
This journal is © The Royal Society of Chemistry 2015
Energy Environ. Sci.
,2015,
8
,16
25 |
23
Opinion
Energy & Environmental Science
Published on 19 September 2014. Downloaded on 15/01/2015 15:34:38.
View Article Online
14 G. Hodes, L. Thompson, J. DuBow and K. Rajeshwar,
J. Am.
Chem. Soc.
, 1983,
105
, 324
330.
15 B. D. James, G. N. Baum, J. Perez and K. N. Baum,
Technoeconomic Analysis of Photoelectrochemical (PEC)
Hydrogen Production
, 2009.
16 A. J. Bard and M. S. Wrighton,
J. Electrochem. Soc.
, 1977,
124
,
1706
1710.
17 R. Nou
, A. J. Frank and A. J. Nozik,
J. Am. Chem. Soc.
, 1981,
103
, 1849
1850.
18 M. G. Walter, E. L. Warren, J. R. McKone, S. W. Boettcher,
Q. Mi, E. A. Santori and N. S. Lewis,
Chem. Rev.
, 2010,
110
,
6446
6473.
19 M. Heyrovsky and R. G. W. Norrish,
Nature
, 1963,
200
, 880
881.
20 W. J. Albery,
Acc. Chem. Res.
, 1982,
15
, 142
148.
21 B. Parkinson,
Acc. Chem. Res.
, 1984,
17
, 431
437.
22 In accordance with ref. 4 and 8, we have chosen to use the
term
regenerative photoelectrochemical cell
to denote a
cell that utilizes a solid/electrolyte junction to generate
electricity, with no net chemical conversion, under
illumination. This usage is in contrast to that of ref. 7,
which refers to these systems as
photovoltaic
devices. As
described above, we have reserved the term
photovoltaic
to denote a system that utilizes a junction between two
electronic conductors, in accord with broad usage and
with the de
nition in many dictionaries of the term
photovoltaic
.
23 In accordance with ref. 24, we have used the term
photoelectrosynthetic
to encompass all devices in
which a net endergonic chemical conversion takes place.
We recognize that the term
photoelectrolytic
has been
used (
e.g.
, ref. 7) in this context to refer to devices in
which a net chemical conversionoccurs with
D
G
>0,and
the term
photocatalytic
has been used to refer to
devices with
D
G
< 0. We chose our nomenclature to
avoid confusion with the common usage of the term
photocatalyst
.
24
Photochemical conversion and storage of solar energy
, ed. J.
Connolly, Academic Press, New York, 1981.
25 C. M. Gronet, N. S. Lewis, G. Cogan and J. Gibbons,
Proc.
Natl. Acad. Sci. U. S. A.
, 1983,
80
, 1152
1156.
26 M. Gr
̈
atzel,
J. Photochem. Photobiol., C
, 2003,
4
, 145
153.
27 M. S. Wrighton, A. B. Ellis, P. T. Wolczanski, D. L. Morse,
H. B. Abrahamson and D. S. Ginley,
J. Am. Chem. Soc.
,
1976,
98
, 2774
2779.
28 C. Levy-Clement, A. Heller, W. A. Bonner and
B. A. Parkinson,
J. Electrochem. Soc.
, 1982,
129
, 1701
1705.
29 L. Alibabaei, M. K. Brennaman, M. R. Norris, B. Kalanyan,
W. Song, M. D. Losego, J. J. Concepcion, R. A. Binstead,
G. N. Parsons and T. J. Meyer,
Proc. Natl. Acad. Sci. U. S. A.
,
2013,
110
, 20008
20013.
30 J. R. Swierk and T. E. Mallouk,
Chem. Soc. Rev.
, 2013,
42
,
2357
2387.
31 K. J. Young, L. A. Martini, R. L. Milot, R. C. Snoeberger III,
V. S. Batista, C. A. Schmuttenmaer, R. H. Crabtree and
G. W. Brudvig,
Coord. Chem. Rev.
, 2012,
256
, 2503
2520.
32 J. Brillet, J.-H. Yum, M. Cornuz, T. Hisatomi, R. Solarska,
J. Augustynski, M. Graetzel and K. Sivula,
Nat. Photonics
,
2012,
6
, 824
828.
33 R. van de Krol, Y. Liang and J. Schoonman,
J. Mater. Chem.
,
2008,
18
, 2311
2320.
34 H.-J. Lewerenz and L. Peter,
Photoelectrochemical Water
Splitting
, The Royal Society of Chemistry, 2013.
35 F. Lin and S. W. Boettcher,
Nat. Mater.
, 2014,
13
,81
86.
36 E. Aharon-Shalom and A. Heller,
J. Electrochem. Soc.
, 1982,
129
, 2865
2866.
37 R. N. Dominey, N. S. Lewis, J. A. Bruce, D. C. Bookbinder and
M. S. Wrighton,
J. Am. Chem. Soc.
, 1982,
104
, 467
482.
38 N. C. Strandwitz, D. J. Comstock, R. L. Grimm, A. C. Nichols-
Nielander, J. Elam and N. S. Lewis,
J. Phys. Chem. C
, 2013,
117
, 4931
4936.
39 A. C. Nielander, M. J. Bierman, N. Petrone, N. C. Strandwitz,
S. Ardo, F. Yang, J. Hone and N. S. Lewis,
J. Am. Chem. Soc.
,
2013,
135
, 17246
17249.
40 M. Barroso, A. J. Cowan, S. R. Pendlebury, M. Gr
̈
atzel,
D. R. Klug and J. R. Durrant,
J. Am. Chem. Soc.
, 2011,
133
,
14868
14871.
41 M. Barroso, C. A. Mesa, S. R. Pendlebury, A. J. Cowan,
T. Hisatomi, K. Sivula, M. Gr
̈
atzel, D. R. Klug and
J. R. Durrant,
Proc. Natl. Acad. Sci. U. S. A.
, 2012,
109
,
15640
15645.
42 M. Barroso, S. R. Pendlebury, A. J. Cowan and J. R. Durrant,
Chem. Sci.
, 2013,
4
, 2724
2734.
43 A. Heller, in
Photoe
ff
ects at Semiconductor-Electrolyte
Interfaces
, American Chemical Society, 1981, vol. 146, ch. 4,
pp. 57
77.
44 A. Heller,
Acc. Chem. Res.
, 1981,
14
, 154
162.
45 S. R. Lunt, L. G. Casagrande, B. J. Tu
s and N. S. Lewis,
J.
Phys. Chem.
, 1988,
92
, 5766
5770.
46
Basic Research Needs for Solar Energy Utilization, US
Department of Energy
, ed. N. S. Lewis and G. Crabtree,
O
ffi
ce of Science, Washington, DC, 2005.
47 M. F. Weber and M. J. Dignam,
J. Electrochem. Soc.
, 1984,
131
, 1258
1265.
48 S. Hu, C. Xiang, S. Haussener, A. D. Berger and N. S. Lewis,
Energy Environ. Sci.
, 2013,
6
, 2984
2993.
49 M. F. Weber and M. J. Dignam,
Int. J. Hydrogen Energy
, 1986,
11
, 225
232.
50 Although the Turner Cell (ref. 53) is classi
ed as a PV-biased
PEC cell, the surface of the p-GaInP
2
at the p-GaInP
2
/3M
H
2
SO
4
(aq) interface is covered in a thin layer of Pt. This
suggests the possibility that the
I
V
characteristics of this
device may be independent of the nature of the electrolyte
at the p-GaInP
2
/3M H
2
SO
4
(aq) interface, which would then
result in the appropriate designation of this cell as a PV-
biased electrosynthetic cell. However, as noted in ref. 35,
the
I
V
behavior of electrodes consisting of thin layers of
Pt on semiconductors is o
en dependent on the pH of the
electrolyte. Thus, based on the available data, is not clear
whether the Turner Cell is more appropriately designated
as PV-biased photoelectrosynthetic cell or as a PV-biased
electrosynthetic cell. A similar argument is relavent for the
devices discussed in ref. 51.
24
|
Energy Environ. Sci.
,2015,
8
,16
25
This journal is © The Royal Society of Chemistry 2015
Energy & Environmental Science
Opinion
Published on 19 September 2014. Downloaded on 15/01/2015 15:34:38.
View Article Online
51 F. F. Abdi, L. Han, A. H. M. Smets, M. Zeman, B. Dam and
R. van de Krol,
Nat. Commun.
, 2013,
4
, 2915.
52 A. J. Bard and T. E. Mallouk,
J. Phys. Chem.
, 1993,
97
, 7127
7128.
53 O. Khaselev and J. A. Turner,
Science
, 1998,
280
, 425
427.
54 S. C. Kondapaneni, D. Singh and O. N. Srivastava,
J. Phys.
Chem.
, 1992,
96
, 8094
8099.
55 K. Domen, S. Naito, M. Soma, T. Onishi and K. Tamaru,
J.
Chem. Soc., Chem. Commun.
, 1980, 543
544, DOI: 10.1039/
c39800000543.
56 We have used the term
photocatalyst
to describe these
systems, in accordance with common parlance and in
accord with the usage of ref. 57. We recognize that
photocatalysis
is o
en, however, used to refer speci
cally
to the light-driven catalysis of a reaction for which
D
G
<
0 (ref. 7).
57 T. Hisatomi, J. Kubota and K. Domen,
Chem. Soc. Rev.
, 2014,
DOI: 10.1039/c3cs60378d.
58 K. R. Gopidas, M. Bohorquez and P. V. Kamat,
J. Phys. Chem.
,
1990,
94
, 6435
6440.
59 N. Serpone, E. Borgarello and M. Gratzel,
J. Chem. Soc.,
Chem. Commun.
, 1984, 342
344, DOI: 10.1039/
c39840000342.
60 J. Weiss and D. Porret,
Nature
, 1937,
139
, 1019
1020.
61 D. Gust, T. A. Moore and A. L. Moore,
Acc. Chem. Res.
, 2009,
42
, 1890
1898.
62 S. C. Silver, J. Niklas, P. Du, O. G. Poluektov, D. M. Tiede and
L. M. Utschig,
J. Am. Chem. Soc.
, 2013,
135
, 13246
13249.
63 L. M. Utschig, S. C. Silver, K. L. Mulfort and D. M. Tiede,
J.
Am. Chem. Soc.
, 2011,
133
, 16334
16337.
64 P. Du, J. Schneider, G. Luo, W. W. Brennessel and
R. Eisenberg,
Inorg. Chem.
, 2009,
48
, 4952
4962.
65 B. D. Yuhas, A. L. Smeigh, A. P. Douvalis, M. R. Wasielewski
and M. G. Kanatzidis,
J. Am. Chem. Soc.
, 2012,
134
, 10353
10356.
66 J. R. Bolton, S. J. Strickler and J. S. Connolly,
Nature
, 1985,
316
, 495
500.
67 B. M. Kayes, H. Nie, R. Twist, S. G. Spruytte, F. Reinhardt,
I. C. Kizilyalli and G. S. Higashi,
Presented in part at the
37th IEEE Photovoltaic Specialists Conference
, Seattle, WA,
2011.
68 C. H. Henry,
J. Appl. Phys.
, 1980,
51
, 4494
4500.
69 E. L. Warren, J. R. McKone, H. A. Atwater, H. B. Gray and
N. S. Lewis,
Energy Environ. Sci.
, 2012,
5
, 9653
9661.
This journal is © The Royal Society of Chemistry 2015
Energy Environ. Sci.
,2015,
8
,16
25 |
25
Opinion
Energy & Environmental Science
Published on 19 September 2014. Downloaded on 15/01/2015 15:34:38.
View Article Online