The growing demand for sustainable lithium production necessitates the development of efficient and environmentally responsible technologies for impurity removal from complex brines. This study presents a novel dual-extraction system capable of simultaneously removing two major contaminants—magnesium and boron—which significantly hinder lithium recovery. Unlike conventional unimetallic approaches that target individual impurities, our method demonstrates high selectivity and efficiency for both elements. A key innovation is the use of p-cymene as a green organic solvent, replacing kerosene-based mixtures while maintaining high extraction efficiencies (>89 % for Mg and 100 % for B) and minimal lithium co-extraction (13 %). To optimize this multivariable system, we employed the Draper-Lin Design (DLD), which enabled robust statistical modeling with only 24 experiments—representing a 33–52 % reduction compared to classical methods such as Central Composite Design (CCD) or Response Surface Methodology (RSM) by Box-Behnken. DLD not only minimized experimental burden but also revealed critical nonlinear interactions and trade-offs between variables, including the dual effect of temperature and the strong interdependence between pH and extractant concentration. These insights allowed the identification of precise operational windows, contributing to the overall robustness and scalability of the process. The proposed system offers important operational and environmental benefits by reducing chemical waste, energy consumption, and processing complexity. These advantages, coupled with the diagnostic power of the DLD approach, position this technology as a promising candidate for industrial application in the sustainable extraction of lithium from high-salinity brines.