nature photonics | VOL 3 | JULY 2009 | www.nature.com/naturephotonics 373
news & views
At present, most commercial photovoltaic devices are made from inorganic semiconductors, such as
silicon, and operate with efficiencies of
around 10–20%. Any significant increases
in efficiency, achieved at low cost, would
clearly accelerate the acceptance of solar
energy as a mainstream renewable energy
source. Organic solar cells represent an
important opportunity to reduce the
manufacturing cost of photovoltaic panels,
if their efficiency can be improved.
On page 406 of this issue, Hardin et al.
report an increase in the range of photon
energies that can be efficiently harvested
by organic solar cells1. The key to the
improvement is a relay dye that absorbs
energy at green wavelengths and transfers it
by Förster resonant energy transfer (FRET)
to a conventional sensitizer dye that
absorbs at red wavelengths. The cell shows
an improvement in conversion efficiency
of almost a third, compared with a cell
without the relay dye.
The main difficulty in improving
the efficiency of photovoltaic energy
conversion is the spectral bandwidth of
sunlight. In conventional inorganic solar
cells, the use of suitable bandgap materials
allows absorption from the near-infrared to
the ultraviolet. In contrast, dye-sensitized
solar cells rely on molecular transitions
where panchromatic absorption is much
harder to achieve.
Conventional dye-sensitized solar cells
are composed of dye molecules that are
anchored to a nanoporous titania film.
After absorption of photons in the dye, an
electron is transferred from the dye into
the titania film and then to an electrical
circuit, where it delivers energy to a load
(Fig. 1a). The electron is subsequently
returned to the ground state of the dye
via a liquid electrolyte that fills the pores
in the nanoporous titania film. Much
effort has been devoted to developing
molecular dyes capable of absorbing the
visible, and some of the near-infrared,
components of the solar spectrum2,3. With
the appropriate choice of dye, wavelengths
of 400–700 nm can be absorbed,
resulting in power-conversion efficiencies
of up to 11.3% (ref. 4).
One way to relieve the difficulties of
achieving panchromatic absorption with a
single dye is co-sensitization with multiple
dye species5. This approach, however, is
limited (as with any surface) by the fact that
different dye species compete for a finite
number of anchor points on the titania
film surface, where electrons transfer
from the dye to the electrical circuit. Thus,
although the spectral range for absorption
can be easily increased by co-sensitization,
increasing the overall efficiency is much
more difficult.
Hardin et al. demonstrate an alternative
approach to co-sensitization that extends
the spectral range of absorption while
preserving the overall optical density
of the dye-sensitized solar cell. They
achieve this by adding a second dye — the
relay — which relies on FRET to transfer
the excited state from the relay dye to the
sensitizer dye (Fig. 1b), into the electrolyte
of a dye-sensitized solar cell. The relay dye
does not compete for the anchor points on
the titania surface and, in addition, absorbs
photons of a different wavelength.
Hardin et al. fabricated cells using a
phthalocyanine sensitizer, chloroform
electrolyte solvent and a perylene-based
relay dye (PTCDI). In this schema,
the sensitizer dye absorbs light with a
wavelength of 600–700 nm, and the relay
dye in the range of 500–600 nm. Overall, a
26% increase in conversion efficiency was
found for the cell containing the relay dye
(compared with a control cell where no
dye was present). It was shown that under
appropriate conditions, FRET can become
the dominant relaxation process for the
relay dye, overcoming the rapid quenching
that one might expect from the presence of
the electrolyte.
The use of energy-relay dyes enables
the broad solar spectrum to be absorbed
by multiple dyes, thereby easing the design
constraints for the dye. However, efficient
FRET-coupling requires that the relay dye
absorbs photons at a higher energy than
that of the sensitizer dye. As a result, this
technique can be used to improve the
short-wavelength photovoltaic response of
the cell, but cannot be used to extend the
range into the near infrared. In principle,
however, the absorption wavelength of
the sensitizer dye can be moved further
into the infra-red, allowing the relay
dye to absorb a greater fraction of the
visible spectrum.
Another constraint is that the dye
must be soluble in the electrolyte, which
places limitations on the electrolyte’s
composition. In Hardin’s study, this results
in an overall reduction of the solar-cell
efficiency. Despite this, an increased
power-conversion efficiency using the relay
dye is observed when compared with a
PHOTOVOLTAIC TECHNOLOGY
Relay dye boosts efficiency
Resonant energy-relay between two dye species allows photovoltaic harvesting of photons across a wider spectral
range. This technique has been exploited to boost the efficiency of dye-sensitized solar cells by 26%.
Daniel J. Farrell and Nicholas J. Ekins-Daukes
Sensitizer
dye
Load
Relay dyeSensitizerdye Titania Titania
Load
Energy
transfer
a b
Figure 1 | Schematic operating principles for a conventional dye-sensitized solar cell where a single dye
must absorb a broad range of wavelengths (a), and an energy-relay dye-assisted solar cell where the
absorption of sunlight light is shared between two molecular species (b).
nphoton_N&V_JUL09.indd 373 16/6/09 14:10:12
© 2009 Macmillan Publishers Limited. All rights reserved