Discovery of novel phosphors for use in light emitting diodes using heuristics optimization-assisted combinatorial chemistry

Abstract

Non-dominated sorting genetic algorithm (NSGA) and particle swarm optimization (PSO), the most well-known population-based heuristics optimization techniques, were used along with high-throughput synthesis and a characterization technique to discover a new alkali–alkaline earth-phosphate phosphor in a multi-compositional search space that consisted of 10 oxides: Li₂O, K₂O, Na₂O, SrO, CaO, BaO, MgO, ZnO, P₂O₅, and Eu₂O₃. Because the number of possible combinatorial candidates in this search space was infinite, a simple high-throughput process would not have accomplished an effective screening. We employed NSGA for a preliminary screening in terms of luminescence intensities at excitations of 254 and 390 nm. Following the NSGA implementation, a fine-tuning process was implemented using PSO in a reduced composition space. In contrast to model-based simulations, we used high-throughput experimentation (HTE) for the experimental evaluation of objective functions in NSGA and PSO executions. As a result, we pinpointed a promising green-light-emitting phosphor, K_0.2188Sr_0.0045Ba_0.1079Ca_0.3647P_0.2977O_δ:Eu_0.0065, exhibiting a monoclinic structure with unit cell parameters, a = 13.2229(1), b = 9.40612(1), c = 12.9628(1), β = 106.765, in the P2₁/a space group, which has potential for use in white light emitting diodes (LEDs).