1.3.1. Linear optical spectroscopy – a starting point for nonlinear spectroscopy The fundamental processes involved in the operation of organic photovoltaics (OPV) can be investigated using the interaction of light with matter. For instance, the linear absorption spectrum contains information about this interaction of light with matter in a thermal equilibrium state. In other words, linear spectroscopy is a useful tool to investigate the available transitions of a material from the ground state to an excited state. For instance, the absorption edge indicates the optical bandgap between the HOMO and LUMO. The presence of several absorption peaks represents vibronic transitions or transitions to higher energy levels, such as from HOMO to LUMO+1. Materials that can form nanocrystals have an additional shoulder of absorption at low energies around the bandgap due to the HOMO-LUMO energy splitting38. Moreover, the so-called ground state charge transfer complexes (CTCs) can be identified by the renormalization of the HOMO and LUMO levels of the whole system86-88. However, for accessing the changes in optical properties induced by light absorption, one should move to nonlinear spectroscopy.
1.3.2. Nonlinear spectroscopy
A nonlinear spectroscopy is capable of disclosing the state of the material when it is not in the equilibrium state (ground state). Amid the popular nonlinear spectroscopy methods are photoluminescence measurements, which are based on the light absorption by a material and subsequent luminescence of the photoexcited state. Another group of popular measurements is based on the photoexcitation of the material and subsequent detection of the change of transmission of light, e.g. pump-probe spectroscopy. In the following sections, a brief overlook will be provided on the light-induced processes and their spectroscopic signatures.
1.3.3. Singlet excitons and their spectroscopic signatures
Typical primary photoexcitations in organic materials for solar cells are singlet excitons that have a net spin of 0 (fig. 1.2). Singlet excitons may be detected by emission if the quantum yield is high enough. The detection of emission is a popular method48-52 to characterize pristine donor and acceptor materials separately and in some cases their BHJs.
Another signature of singlet excitons is their absorption from the lowest excited state to
16
higher energy levels, e.g. S1 to S2, which has lower energy than the bandgap of the material (fig. 1.2, example spectrum is shown in fig. 1.10). Singlet excitons can also be detected by the decreased probability of the S0 to S1 transitions due to the depletion of the ground state (called Ground State Bleaching, GSB).
Fig. 1.10 Example of typical IR PIA spectra of organic photovoltaic materials.
The formation of a BHJ is usually detrimental for the emission yield due to an efficient exciton dissociation rendering a very low signal to noise (e.g.<< 1) in photoluminescence measurements. The drawback of detecting transitions from S1 to the higher energy levels for spectroscopy is that these energies often overlap with other available transitions. For instance, charge induced transitions89 often have some overlap with singlet transitions. The GSB is present when the material is not in the ground state, therefore, it cannot be attributed to singlet excitons alone.
1.3.4. Charges: polarons and their spectroscopic signatures
The LE and HE charge-induced (polaronic) transitions have already been introduced in Section 1.1.4 (and fig. 1.5). However, LE and HE transitions are only a simplified representation. The shifted molecular orbitals form intrinsically broad photoinduced absorption bands (PIA) because electronic transitions are mixed with vibrational levels (example spectra are shown in fig. 1.10). These multiple charge-induced (polaronic)
LE HE
Exciton
Charge (Polaron)
CT state Charge (Polaron)
17 transitions forming PIA bands are called the LE and HE polaron bands. Due to the very broad nature of polaron bands, the HE polaron band usually overlaps with PIA of excitonic transitions, whereas LE polaron band often has overlap with PIA of transitions from CT states.
1.3.5. Spectral positions of excited states in typical materials for organic photovoltaics
Typical absorption spectra of organic donor materials for photovoltaics span the range of visible to near-infrared (NIR). Emission by singlets is red-shifted with respect to absorption due to the vibrational relaxation. Further absorption by the excitons to their higher energy levels occurs at spectral energies lower than the original bandgap of the material.
These energies often overlap with the emission spectra. The charge-induced (polaron-induced) HE absorption is also below the bandgap of the material, usually in the NIR region.
The LE charge-induced absorption, on the other hand, is situated more in the middle-IR (Mid-IR) range. Therefore, the LE band is advantageous for selective detection of charges.
However, in some combinations of materials transitions related to CT states (or CT excitons) may overlap with the LE polaron absorption86-88, 90,91
(see example spectra in fig. 1.10), which requires spectral decomposition. In principle, CT states and CS states (separated charges) should have slightly different absorption spectra due to the difference of molecular geometry or deformation of the molecules to the different extent. However, discrimination between them may not be a trivial task due to the broad spectral response of polaron bands. In contrast, when LE transitions have no overlap with the transitions related to CT states, the data analysis becomes more straightforward as compared to, for instance, probing of the HE polaron band.
1.3.6. Concept of the pump-probe technique
In the presence of any excited state, be it singlet or triplet excitons or charges, all of them induce additional optical transitions in a spectral energy range below the bandgap energy. Simultaneously, the ground state is bleached and emission from the excited state may occur. One of the most popular methods to follow the time evolution of different excited states is the so-called pump-probe technique. Time-resolved pump-probe spectroscopy is capable of addressing the transient phenomena.
The basic concept of TR pump-probe experiments is that one laser pulse, called excitation or pump pulse, creates a population of excited states while a second, time-delayed
18
probe pulse detects the changes induced by the pump pulse (fig. 1.11). Changing the delay between the two pulses allows monitoring the dynamics of excited states. The shorter the laser pulses, the faster the processes that can be monitored. The time resolution of the processes that can be resolved is limited by the combination of the signal-to-noise ratio and the temporal width of the convolution of the pump and probe pulses, defined as the full width at half maximum (FWHM) or standard deviation σ of the time profile of the measured response, where for Gaussian profiles, typical for laser pulses.
Fig. 1.11 Schematic representation of a pump-probe setup. The pump pulse induces changes in the transmission ΔT as imposed by the chopper, which are detected by the time-delayed probe pulse.
1.3.7. Polarization-sensitive detection: photoinduced transient anisotropy
Another useful approach based on the pump-probe scheme is to monitor the polarization of the transition dipole moment of excited states. This technique allows getting information that is otherwise hidden in the simple pump-probe technique, where only the population of excited states is monitored.
The concept of polarization sensitive technique is that correlations are measured between the two transient dipole moments of the transition to the excited state induced by excitation pulse and from the excited state induced by the probe pulse (fig. 1.12). This concept is realized in the following way: the pump pulse that provides a photoexcitation is linearly polarized in the vertical direction (fig. 1.12). The probe pulse has its polarization rotated by 45 ° with respect to the pump pulse. After the probe pulse passes the sample, it is split into two replicas. Polarizers are placed in front of the detectors in order to select the components of both replicas that are parallel and perpendicular to the pump pulse. The