In this dissertation, we describe our work into suppressing the formation of triplet excitons in organic solar cells, created as a result of unavoidable non-geminate charge recombination processes. Triplet excitons represent a major non-radiative loss pathway in organic solar cells and through this work, we demonstrate tactics to reduce the non-radiative decay of photo-generated charges and thereby enhance the radiative efficiency of their decay processes. As the ideal solar cell would have only voltage loss resulting from the radiative recombination of charge carriers, we aim to decrease the magnitude of the non-radiative voltage losses in organic solar cells. In this work, we present two distinct strategies to achieve this: i. The kinetic suppression of the back electron transfer process from the triplet charge transfer state to the local triplet state, so it is out-competed by re-dissociation of the charge transfer state into free charges. ii. The use of low-exchange energy materials that allow for the simultaneous creation of a charge transfer state that is close in energy to the lowest singlet state to minimise the energy loss associated with charge generation, but still below the energy of the lowest triplet state, thermodynamically forbidding the formation of any local triplet states. The first study investigates two closely-related high-performing organic solar cells that show an extremely low total voltage loss. Through detailed investigations of these donor-acceptor blends, we determine that the low voltage loss can be attributed to enhanced levels of radiative recombination. Further, we show that over the timescales of non-geminate recombination there is no observable formation of triplet excitons in the blend. Accrediting the enhanced radiative recombination to the absence of the non-radiative triplet loss pathway, we explore the factors controlling the rate of the back electron transfer process that leads to triplet formation. From our preliminary calculations, we determine that the factor most likely responsible for a slow back electron transfer rate is the electronic coupling between the triplet charge transfer state and the molecular triplet exciton. If the rate of this process is sufficiently slowed, it will be out-competed by the rate of the re-dissociation of the charge transfer state, leading to the kinetic suppression of triplet exciton formation. Having determined the efficacy of removing the triplet loss pathway in enhancing the radiative efficiency of recombination, the next two investigations focus on the second strategy discussed: thermodynamically forbidding triplet formation. For this, we utilise two different low exchange energy organic thermally activated delayed fluorescence materials as electron acceptors that are paired with wider band gap, high triplet energy donors. Through spectroscopic studies of the blends, we demonstrate that charge generation can proceed efficiently from both components, a prerequisite for efficient solar cell operation. Furthermore, the excellent radiative efficiencies and low non-radiative voltage losses of the blends confirms that this is indeed an effective route to improve the performance of organic solar cells.