Singlet Exciton Fission

Although the possibility of singlet exciton fission and triplet exciton fusion was recognized early on, just a few years after the invention of the laser, the fundamental physical processes behind these effects are currently being studied with a renewed focus, in part thanks to the availability of new technology, and also because of the general interest of multi-exciton generation for a more efficient harvesting of higher energy photons in photovoltaics.

A simple description of singlet fission starts from a pair of one ground-state and one excited-state molecule. What happens next can be described by the following somewhat cartoonish depiction.

A singlet exciton is represented in a molecular picture by a pair of molecules, one in the ground state, one in the first excited state. An initial fission step leads to two molecules, both in the first electronic excited state, but each of them in a triplet state, which can lower the energy stored in each molecule to equal or less than half the energy that was stored in the molecule in a singlet excited state. The pair of triplet excitons created by fission must be in an entangled spin-state, with a total spin of zero. After this initial step, the triplets in the triplet pair can relax, move away from each other, loose their entanglement, and become completely independent. We don't know yet when and how all these processes happen.

All intermediate triplet-pair states can in general (depending on relative singlet and triplet energies) undergo geminate fusion back into a singlet state if they have enough energy, allowing for photon emission (this is the case in rubrene). But the most probable outcome in a material like rubrene is a completed fission process that results in two separate triplet excitons that independently diffuse in the crystal.

Much of the research on the initial step in the fission process, when the photoexcited singlet state first transitions to a triplet-pair, was done by detecting the disappearance and/or the appearance of absorption bands assigned to singlet or triplet states, respectively. The fission times reported in the literature, however have vary from femtoseconds to picoseconds, and there is the possibility that variations in excitation conditions may affect the results. As an example, using pulses much shorter than 100 fs makes the excitation energy uncertain, and short pulses in general has such high peak intensities that multiphoton excitation becomes possible or even probable. Both issues could in principle affect the results for the rate of singlet fission.

Two laser beams create a photoexcitation grating, a third beam detects what happens because the grating diffracts it to create the red output beam.

We have started an investigation of triplet fission that relies on photoexcitation with low intensity 1 ps laser pulses in a transient grating configuration. In this way we can exploit the intrinsic sensitivity of forced light-scattering in a grating to study the fission system under lower perturbation, and at lower repetition rates.

For the first investigations, we exploited the expected ability of triplet excitons to be photoexcited to an higher excited state both by photons near an energy of 1.6 eV and also by photons near an energy of 2.4 eV, which is also the energy range that allows photoexcitation of singlet states from the ground states. In this way it is possible ot use photons near an energy of 2.4 eV to both photoexcite singlet states and detect triplet states, while at the same time the 2.4 eV photon is not absorbed by a photoexcited singlet state. Tuning the wavelengths of the pulses allows to match the energy of a photon to the main peaks of rubrene's vibronic progression. But this is a pump and probe experiment where two pump pulses photoexcite a spatially sinusoidal distribution of singlet states, and a third pump pulse, arriving later, detects the evolution of the corresponding sinusoidal distribution of triplet states that are created by fission. We started doing this experiment using the same photon energy for excitation and detection.

Two laser beams create a photoexcitation grating, a third beam detects what happens because the grating diffracts it to create the red output beam.

Time evolution of the the grating amplitude detected while the photon energy was varied between 2.3 and 28 eV.

The data obtained until now shows a slower build-up of the detected diffraction signal as the energy of the photons is increased. If the probe pulse is only sensitive to triplet states because of its wavelength, then this means that triplet states are generated more slowly if the original singlet state is excited into a higher vibrational levels. But it is also possible that, thanks to the sensitivity of grating diffraction, the probe pulse detects more than just the increase in triplet population. Experiments are ongoing to investigate this issue.