Singlet Fission - Department of Chemistry

D Chemical Reviews, XXXX, Vol. xxx, No. xx Smith and Michl
Figure 3. Dye-sensitized solar cell that uses a singlet fission
sensitizer (C1) in conjunction with a conventional sensitizer (C2).
The C1 sensitizer comprises the top layer of the cell and absorbs
light above 2 eV. The C2 sensitizer absorbs the remaining light
between 1 and 2 eV. CB, conduction band; VB, valence band; CE,
counterelectrode; NHE, normal hydrogen electrode.
Figure 4. Schematic sketch of theoretical efficiency as a function
of the S0-T1 band gap for a singlet fission solar cell defined
in Figure 3 (red) and a conventional dye-sensitized solar cell
(blue).
carriers separately and quantitatively (Figure 3). The ideal
efficiency of such a singlet fission cell is shown schematically
as a function of the S0-T1 band gap in Figure 4, assuming
a 200% yield of charge carrier pairs per photon. The optimal
location of the S0-S1 absorption edge is ∼2 eV, with the
triplet energy at ∼1 eV, but minor deviations from these
values would have little effect. In the derivation of the
optimum efficiency value close to 1/2, it was assumed that
the layer doped with singlet-fission capable sensitizer absorbing
at 2 eV and above would be immediately followed by a
layer of an ordinary sensitizer capable of absorbing photons
of energies between 1 and 2 eV and generating a single
electron and hole per photon. It is then arguable whether
the assembly can still be called a single junction cell, but
because no current matching is necessary, even if it is viewed
as a double junction cell, it would be one of a particularly
simple kind.
There would be an intrinsic advantage to the use of triplets
for electron or hole injection, since back electron transfer to
yield the ground state would be spin-forbidden. The longer
lifetime of the triplet would provide more opportunity for
charge separation but also more opportunity for quenching,
for example, by charge carriers.
A sensitizer capable of singlet fission needs to contain
more than one excitation site. While the use of covalent
dimers or oligomers in dye sensitized solar cells is similar
to the already well established use of individual molecular
dyes, 27-29 the injection of charges from molecular crystals,
aggregates, or polymers is likely to present challenges. This
is at least partially due to the fact that single molecules can
be more easily specifically and covalently bound to oxide
surfaces, securing efficient electron transfer and charge
collection. The difficulty with crystals is that they would need
to be extremely small in order to be useful as semiconductor
sensitizers. Nanocrystalline semiconductor films used in
photovoltaic cells, such as TiO2, typically have a porosity
of ∼50% and a grain size in the range of 10-80 nm, 29 and
charge transfer to the macroscopic electrode occurs by grainto-grain
hopping. For efficient charge injection into the
semiconductor and its further transfer to electrode, the
sensitizer nanocrystals probably should be significantly
smaller than the semiconductor grains on which they are
adsorbed. In this respect, work with flat or bulk heterojunctions
may be easier.
At present, the proposed use of singlet fission is at the
stage of fundamental research. Even if one disregards entirely
all possible practical problems that have not been addressed,
such as cost and long-term stability in sunlight, basic
molecular engineering problems remain. These challenges
include matching energy levels appropriately for fast yet lowpotential-loss
charge separation and transfer, and preventing
premature charge injection from the short-lived originally
excited singlet state while assuring efficient charge injection
from the long-lived triplet excited state.
At an even more fundamental level, it appears sensible to
divide the problem of finding a useful singlet fission system
into three parts. (i) Which chromophore should be used to
maximize the rate of singlet fission while minimizing the rate
of the reverse process and also meeting other conditions such
as a high absorption coefficient at all energies above the onset
of absorption, which needs to be located near 2 eV? (ii) How
should neighboring chromophores be coupled into pairs or
higher aggregates? Relatively strong coupling is needed if
singlet fission is to be very fast. (iii) How should we ensure
independent behavior of the two resulting triplets that would
permit them to undergo two independent charge separation
steps, avoiding the danger that the hole (or electron) left behind
after the first injection step will quench the remaining triplet
before the second injection can take place? The quenching of
triplets by spin 1/2 particles is well known and tends to be fast.
Relatively weak coupling is obviously needed.
Assuring a coupling that is simultaneously strong to ensure
fast singlet fission and weak to allow the triplets to move
apart or act as if they were far apart sounds like magic.
However, although it has not been demonstrated, it appears
that in this case there is a chance that one may actually be
able to have one’s cake and eat it, too (section 2.3).
Work on the problem has barely begun. Of the three
fundamental tasks, the first has received attention in the
literature. 30,31 As a result, some theoretical guidance toward
chromophore structures likely to be efficient in singlet
fission is available (section 2.2.1). Although it has already
been put to use, 32 as will be seen below, very few organic
materials have been examined so far. Little has been
published 33 on the second task, optimization of chromophore
coupling for fast singlet fission (section 2.2.2). At the
moment, neat (pure) solid materials and aggregates are the
only ones for which highly efficient singlet fission with triplet
yields up to 200% has been demonstrated, and they need