Analysis of Photoinduced Carrier Recombination Kinetics in Flat and Mesoporous Lead Perovskite Solar Cells

In this work, we analyze the carrier recombination kinetics and the associated charge carrier density in methylammonium lead iodide perovskite (MAPI) solar cells that use mesoporous TiO2, as selective contact (m-MAPI) and flat solar cells (without the mesoporous TiO2 f-MAPI), which are the most common device architectures for perovskite solar cells. The use of PIT-PV (photoinduced transient photovoltage) and L-TAS (laser transient absorption spectroscopy) showed that for devices that cannot reach efficiencies close to 19% there is a slow component of the photovoltage decay that corresponds to a charge recombination pathway for carrier losses responsible for the lower device efficiency. Moreover, we have also identified a primary interfacial charge recombination pathway for carrier losses that is common in all devices studied, independent of their efficiency or their device structure, which we have associated with the recombination reaction between electrons in the perovskite and holes in the organic semiconductor material used as the selective contact.

Charge collection and pore filling in solid-state dye-sensitized solar cells

Ilkay Cesar, Michael Graetzel, Robin Humphry-Baker, Henry Snaith, Shaik Mohammed Zakeeruddin

The solar to electrical power conversion efficiency for dye-sensitized solar cells (DSCs) incorporating a solid-state organic hole-transporter can be over 5%. However, this is for devices significantly thinner than the optical depth of the active composites and by comparison to the liquid electrolyte based DSCs, which exhibit efficiencies in excess of 10%, more than doubling of this efficiency is clearly attainable if all the steps in the photovoltaic process can be optimized. Two issues are currently being addressed by the field. The first aims at enhancing the electron diffusion length by either reducing the charge recombination or enhancing the charge transport rates. This should enable a larger fraction of photogenerated charges to be collected. The second, though less actively investigated, aims to improve the physical composite formation, which in this instance is the infiltration of mesoporous TiO2 with the organic hole-transporter 2,2',7,7'-tetrakis(N,N-di-p-methoxypheny-amine)-9,9'-spirobifluorene (spiro-MeOTAD). Here, we perform a broad experimental study to elucidate the limiting factors to the solar cell performance. We first investigate the charge transport and recombination in the solid-state dye-sensitized solar cell under realistic working conditions via small perturbation photovoltage and photocurrent decay measurements. From these measurements we deduce that the electron diffusion length near short-circuit is as long as 20 µm. However, at applied biases approaching open-circuit potential under realistic solar conditions, the diffusion length becomes comparable with the film thickness, ~2 µm, illustrating that real losses to open-circuit voltage, fill factor and hence efficiency are occurring due to ineffective charge collection. The long diffusion length near short-circuit, on the other hand, illustrates that another process, separate from ineffective charge collection, is rendering the solar cell less than ideal. We investigate the process of TiO2 mesopore infiltration with spiro-MeOTAD by examining the cross-sectional images of and performing photo-induced absorption spectroscopy on devices with a range of thickness, infiltrated with spiro-MeOTAD with a range of concentrations. We present our interpretation of the mechanism for material infiltration, and by improving the casting conditions demonstrate efficient charge collection through devices of over 7 µm in thickness. This investigation represents an improvement in our understanding of the limiting factors to the dye-sensitized solar cell. However, much work, focused on composite formation and improved kinetic competition, is required to realize the true potential of this concept.

2008

A Material Approach to Dye-Sensitized Solar Cells

Philippe Pierre Labouchère

With growing concerns on future energy supplies, solar energy appears as an energy source whose potential remains to be tapped at a large scale. In the last two decades, dye-sensitized solar cells (DSCs) have been considered as a competitive means to convert solar energy into electricity thanks to their low manufacturing and environmental costs, simple fabrication and high efficiency. Indeed, record efficiencies as high as 13% have been obtained for liquid DSCs and recently, solid-state DSCs using solid hole transport materials as well as perovskite materials as sensitizers have showed efficiencies above 17%. From a physical point of view, charge carrier generation in a DSC takes place in a photoactive light harvester chemisorbed on a mesoscopic semiconductor oxide. Due to the excellent electronic overlap, the exciton separation occurs by the injection of the excited electrons and holes from the dye into the conduction band of the semiconductor and redox electrolyte, respectively. A high surface area mesoscopic photoanode is necessary to ensure a high-density bulk heterojunction and efficient light harvesting in the visible part of the solar spectrum due to effective dye loading. This implies that the DSC has a large, heterogeneous interface through which photogenerated electrons may be intercepted as a result of the slow electron transport. The latter is governed by an ambipolar diffusion mechanism controlled by trap-limited hopping through a long path to the transparent conductive electrode. It is however hindered by the low electron mobility in anatase TiO2 nanoparticles and the multiple grain boundaries in the mesoporous film. The work conducted during this thesis aimed at developing new architectures for the electrodes of DSCs using notably atomic layer deposition (ALD) and inverse opal host-guest structures with different semiconducting metal oxides. First, using a self-assembled polymeric template and ALD, we fabricated a porous ZnO inverse opal backbone. Advantages are numerous: a high electronic mobility of ZnO, a precise control of the metal oxide deposition using ALD and in situ growth of ZnO nanowires at low temperatures. Besides increasing the surface area available for photoactive dye molecules, single-crystal nanowires offer a defect-free path for photoinjected electrons. However, because of the relative instability of ZnO in acidic environments, a TiO2 overlayer was deposited by ALD. This proved to increase the recombination resistance of photogenerated electrons and increased their lifetime in the porous film. This host-guest approach was further transposed to three related fields in photo-electrochemistry using different ALD precursors. For water splitting, a niobium-doped tin oxide inverse opal backbone was constructed, which served as a host for hematite deposited by atmospheric pressure chemical vapour deposition or ALD. In the field of plasmonics, a TiO2 inverse opal host was made to act as a scaffold for silver nanoparticles. For solid-state DSCs, we strived to achieve an ultra-thin ZnO inverse opal, which acted as a host for perovskite crystals. Finally, we studied the effect of very thin TiO2 and ZnO overlayers deposited by ALD on top of mesoporous ZnO and TiO2, respectively. We were able to accurately monitor the thickness of overlayers and study the effects on photovoltaic parameters such as transport and recombination rates of photogenerated electrons, photocurrent, photovoltage, fill factor and efficiency.

EPFL2014

Growth of CH3NH3PbI3 cuboids with controlled size for high-efficiency perovskite solar cells

Michael Graetzel, Jeong-Hyeok Im, Norman Pellet

Perovskite solar cells with submicrometre-thick CH3NH3PbI3 or CH3NH3PbI3-xClx active layers show a power conversion efficiency as high as 15%. However, compared to the best-performing device, the average efficiency was as low as 12%, with a large standard deviation (s.d.). Here, we report perovskite solar cells with an average efficiency exceeding 16% and best efficiency of 17%. This was enabled by the growth of CH3NH3PbI3 cuboids with a controlled size via a two-step spin-coating procedure. Spin-coating of a solution of CH3NH3I with different concentrations follows the spin-coating of PbI2, and the cuboid size of CH3NH3PbI3 is found to strongly depend on the concentration of CH3NH3I. Light-harvesting efficiency and charge-carrier extraction are significantly affected by the cuboid size. Under simulated one-sun illumination, average efficiencies of 16.4% (s.d. +/- 0.35), 16.3% (s.d. +/- 0.44) and 13.5% (s.d. +/- 0.34) are obtained from solutions of CH3NH3I with concentrations of 0.038 M, 0.050 M and 0.063 M, respectively. By controlling the size of the cuboids of CH3NH3PbI3 during their growth, we achieved the best efficiency of 17.01% with a photocurrent density of 21.64 mA cm(-2), open-circuit photovoltage of 1.056 V and fill factor of 0.741.