Enhancing Performance of a Novel Temperature Swing Salting-Out Process for Sustainable Lithium Recovery from Brines

Abstract

Abstract The rapid growth in global lithium demand, together with the limitations of conventional solar evaporation processes, has intensified the need for direct lithium extraction (DLE) technologies with high energy and water efficiency. In this context, the temperature-swing salting-out (TSSO) process, which combines the lower critical solution temperature behavior of the diisopropylamine–water system with salting-out effects, is a promising DLE technology capable of selective lithium enrichment using low-grade heat. However, the sensitivities of key operating factors, their interaction effects, and the nonlinear behaviors governing the TSSO process have not yet been fully elucidated. Accordingly, this study quantitatively identified the key factor effects of the TSSO process and established a performance prediction framework to support rational operating condition selection. A total of 46 experiments were conducted using a rotatable central composite design augmented with four-level uniform design points. Process behavior was systematically analyzed using analysis of variance, Box–Cox-transformed regression modeling, and response surface methodology. The results confirm that the solvent-to-feed ratio is the dominant factor across all performance metrics—lithium recovery, impurity removal efficiency, feed recovery, and lithium enrichment—and distinct trade-off relationships were observed among them. In probabilistic optimization using Monte Carlo simulation, the optimal lithium enrichment point derived from the highest density region of the predictive distribution was experimentally validated, achieving balanced performance values of 541.71% lithium enrichment, 77.38% lithium recovery, and 99.06% impurity removal efficiency. In addition, the operating condition with the highest probability of simultaneously satisfying lithium recovery ≥ 80%, impurity removal efficiency ≥ 99%, and lithium enrichment ≥ 500% was statistically identified. Overall, this study quantitatively elucidates the factor effects of the TSSO process, interprets their behaviors, and identifies probabilistically optimal operating conditions, which collectively enhance the industrial applicability of TSSO as an energy-efficient DLE technology.