Charge carrier dynamics in polymer solar cells : an opto-electronic study

Bulk-heterojunction polymer solar cells are promising alternatives for conventional energy sources. Due to the lower cost of manufacturing, processing and flexibility of polymers, solar cells made of organic materials may compete with current inorganic solar cells. The photoactive layer of bulk-heterojunction solar cells consists of a blend of an electron donating and an electron accepting material forming a nanostructured interpenetrating network. When light is absorbed in the active layer, a photoinduced charge transfer across the interface between donor and acceptor materials can produce a geminate electron–hole pair. This geminate electron–hole pair is expected to be strongly bound by the Coulombic interaction. Due to the low dielectric constant of organic materials, the binding energy may be considerable. Despite the tight binding of the electron and hole at the interface, solar cells with a p-conjugated polymer as donor material and a fullerene as acceptor, show high internal quantum efficiencies. A factor contributing to the high internal quantum efficiency is the long lifetime of the photogenerated charges. Although photoinduced charge transfer from p-conjugated polymers to the fullerene occurs on a subpicosecond time scale, recombination of the photogenerated charges to the ground state is a much slower process. A long lifetime of the photogenerated charges is needed in order to collect the charges in an external circuit before they recombine. Therefore, 0a better understanding of the factors influencing dissociation and lifetimes of the photogenerated charges in polymer photovoltaic devices is necessary. In this thesis the charge recombination dynamics in a bulk-heterojunction polymer–fullerene solar cell has been investigated by a variety of optical, electrical and numerical techniques, presented in Chapters 2 through 6. In Chapter 7 investigation of the photophysics in an all polymer bulk-heterojunction solar cell has been presented. Near-steady-state photoinduced absorption (PIA) spectroscopy was employed to study the charge carrier dynamics in polymer–fullerene (MDMO-PPV:PCBM) solar cells under operating conditions (Chapter 2). It was shown that the electrodes significantly complicate the PIA measurement as compared to the PIA measurement on plain films. Thermoreflection at the aluminium electrode introduces a new absorption band in the PIA spectrum. Reduced absorption of infrared light complicates probing the low energy absorption of the MDMO-PPV radical cation. It was possible to correct for these effects. Reduced absorption of infrared light is an important effect to take into account when designing novel low band gap solar cells. Comparison between photocurrent measurements and PIA measurements under identical conditions showed that the two techniques probe different subsets of the photogenerated charge carriers. Unlike the photocurrent measurements, PIA probes the trapped rather than the non-trapped charge carriers in the device. While nearsteady- state PIA is sensitive to trapped carriers at the microsecond time scale, the photocurrent under identical conditions is mainly due to mobile charge carriers that are not deeply trapped. Electric field modulated PIA (Chapter 3) was used to further investigate the trapped charge carriers in the device. The PIA signal is dominated by trapped charge carriers with lifetimes smaller than the modulation period (14.3 ms). Deeply trapped charge carriers persisting into the ms time scale are few in number and have virtually no influence on the PIA signal and photocurrent. Because the lifetime of trapped carries could not be determined by near-steadystate PIA, the decay of trapped charges was investigated at a much shorter time scale by means of transient non-resonant hole-burning spectroscopy (Chapter 4). In the 10 ns–10 µs time range, the recombination dynamics of the photogenerated charges is limited by the detrapping of the holes in MDMO-PPV. The eventual recombination of the cationic charge carriers with their anionic counterparts takes place over several orders of magnitude in time resulting in a power-law type of decay. This type of decay was explained by a model in which the charge carriers are allowed to hop in an energetically disordered Gaussian distributed density of states of donor and acceptor phases. Monte Carlo simulations (Chapter 5) of dissociation and recombination of an electron–hole pair at a donor–acceptor interface based on this model can reproduce the power-law decay observed experimentally. The simulations show that the dissociation of geminate charge pairs is assisted by disorder, electric field and temperature, and the results can be understood in terms of a two-step model, in which detrapping of the least mobile charge carrier governs the charge recombination kinetics. The recombination dynamics of trapped charge carriers was also studied electrically, by means of time-delayed collection field (TDCF) experiments, presented in Chapter 6. TDCF allowed us to measure in a quantitative way the number of charge carriers present in the device in the time range 1 µs–10 ms, complementing the time-resolved PIA measurements (Chapter 4). Consistent with results presented in Chapter 2 through 4, TDCF shows the existence of long-lived carriers and the powerlaw decay (~t –a) at 80 K. Agreement with PIA measurements is obtained when the influence of background illumination is taken into account. With increasing background illumination, TDCF shows a shortening of the lifetime of the photogenerated charge carriers, which may be attributed to trap filling of the deeper traps by the background illumination. The photophysical properties of a polymer–polymer blend (MDMOPPV: PCNEPV) have been investigated in Chapter 7. In this blend photoexcitation results in a luminescent exciplex. This is direct evidence for formation of bound electron–hole pairs at the donor–acceptor interface. Application of an electric field results in dissociation of the marginally stable exciplex into charge carriers, as expected from MC simulations of the field-assisted dissociation of a bound electron– hole pair (Chapter 5). The singlet exciplex can decay to the triplet exciplex by intersystem crossing and further decay to the lowest triplet state (T1 MDMO-PPV). PIA measurements show an increase of the triplet absorption signal in the blends compared to the pristine materials, because the charge-separated state is higher in energy than the lowest triplet state in the blend. From the perspective of photovoltaic energy conversion, triplet formation in operational solar cells is a disadvantage as it may present a significant loss channel.