Measuring and controlling exciton diffusion, charge generation and charge extraction in organic and hybrid semiconductors for photovoltaic applications

Abstract

The growing demand for energy and the need for renewables as well as the advent of the internet of things increase the demand for more versatile, efficient, cheap and environmentally friendly solar cells based on organic and hybrid semiconductors. Understanding and controlling the underlying processes that govern exciton diffusion, charge generation, charge extraction and domain size is of vital importance to the efficiency, stability and scalability of these devices.

This thesis examines in detail the exciton and charge carrier behaviour in, and the characteristics of new materials for organic and hybrid solar cell applications. A particular focus is put on measuring, understanding and controlling hole extraction from 퐶퐻₃푁퐻₃푃푏퐼₃ to hole extracting layers and overcoming the trade-off between exciton harvesting and charge extraction in small molecule bulk heterojunction organic solar cells. The main method used for this investigation was ultrafast timeresolved spectroscopy, specifically ultrafast optical transient absorption and ultrafast time-resolved fluorescence decay with a streak camera.

A study of hole extraction from 퐶퐻₃푁퐻₃푃푏퐼₃ was carried out for two different hole extracting layers, the standard PEDOT:PSS polymer used in the inverted p-i-n perovskite solar cells and a new nanoparticle NiO low temperature solution processed thin film. The two extraction layers and the 퐶퐻₃푁퐻₃푃푏퐼₃ perovskite active layer were first characterized using optical and physical methods such as UV-Vis spectroscopy and atomic force microscopy as well as air photoemission spectroscopy to confirm that the same perovskite was grown on top of both PEDOT:PSS and NiO and to investigate energy level alignment. A new method based on the ultrafast photoluminescence surface quenching experiment was developed and introduced which allows for the separation of bulk and interfacial effects on charge extraction from thin films by illuminating the samples from opposite sides. This new method was used to compare hole extraction from 퐶퐻₃푁퐻₃푃푏퐼₃ to NiO and PEDOT:PSS. It was found that NiO shows faster hole extraction from the 300 nm thick perovskite film than PEDOT:PSS on the time scale of 300 ps, which is independent of charge carrier density in the region of 10¹⁶-10¹⁷ cm⁻³. The interface with PEDOT:PSS was found to severely limit charge extraction rate at charge densities exceeding 10¹⁶ cm⁻³. Furthermore, the transfer rate was found to decrease with time and to be dependent on charge density in the region 10¹⁶-10¹⁷ cm⁻³ which we interpreted as charge accumulation. These findings were confirmed by transient absorption spectroscopy. Hole diffusion coefficient 퐷 = 2.2 cm²/s ± 0.4 cm²/s and quenching rate k=3.6 × 10⁵ m/s ±0.2 m/s were determined in the perovskite film that were independent of charge density. This indicates a band-like hole transport regime, not observed for solution processed CH₃NH₃PbI₃ films before. Our findings stress the importance of interface optimization in devices based on perovskite active layers as even in the case of the superior quencher, NiO, there is still room for improvement of the interfacial transfer rate.

The trade-off between exciton harvesting and charge extraction in small molecule bulk heterojunction organic solar cells was tackled by employing a post processing method of solvent vapour annealing on thin films of DR3TBDTT:PC₇₁BM and SMPV1:PC₇₁BM. It was found that as a result of annealing with carbon disulfide, the UV-Vis absorption spectrum changes which indicates changes to the structure of the film. It was further revealed, using exciton-exciton annihilation, that the exciton diffusion coefficient and exciton diffusion length are increased (almost 3-fold for the best case) as a result of solvent vapour annealing. Furthermore, enhanced device performance after treatment with carbon disulfide was recorded; this is explained by better charge extraction, an insight revealed by a transient absorption study. Finally, using an all optical method for domain size determination it was found that solvent vapour annealing can be used to increase domain size. Normally increased domain size would have a detrimental effect on device performance due to a loss in the number of excitons which can reach the donor-acceptor interface, but the increased exciton diffusion length allows us to overcome this trade-off and achieve better device performance.

Date of Award	Jun 2020
Original language	English
Awarding Institution	University of St Andrews
Supervisor	Ifor David William Samuel (Supervisor)