Towards sustainable cosmetic Ingredients: A Simplified in silico approach for selecting Innovative natural mixtures as solvents in biomass extraction

• Smart selection strategy allow simple design for cosmetics MINAC. • A user-friendly workflow for selective MINAC based on COSMO RS is provided. • Experimental data validated the new σ-screening approach. • σ-screening was more relevant than activity coefficient in this strategy. • σ-screening will contribute to in silico pre-screening dissemination. The cosmetics industry is under growing pressure to adopt sustainable and efficient methods for biomass extraction. Among emerging solutions, mixtures of natural compounds (MINAC) have recently been recognized as promising, eco-friendly alternatives. To support their development, a Smart Selection strategy has been introduced to guide the design of MINAC systems, ensuring compliance with cosmetic regulations and formulation constraints. Despite this, the experimental screening of suitable formulations remains time-consuming and resource-intensive. To streamline this process, the Conductor-like Screening Model for Real Solvents (COSMO-RS) was applied to further refine a pre-selected library of cosmetics-compatible solvents. This refinement was guided by the nature of the key target compounds to be extracted from the selected biomass. A user-oriented screening methodology is proposed and illustrated through a case study focused on the extraction of fatty acids and non-polar pigments from Spirulina. Within this framework, σ-surfaces, σ-profiles, and σ-potentials were employed to identify solvent mixtures most compatible with the filtered raw materials. This step enabled the construction of a concise solvent library tailored to the target components. Solid–liquid equilibrium (SLE) calculations were then performed to determine the eutectic composition of each preselected MINAC. Four eutectic systems based on caprylic acid were identified as particularly relevant, enabling the a priori selective extraction of saturated fatty acids (SFA), combinations of SFA and polyunsaturated fatty acids (PUFA), or both pigments and PUFA. This method was benchmarked against the conventional activity coefficient approach, commonly regarded as the gold standard. While both in silico -based methods showed overall consistency, the applied filter-based strategy offered practical advantages in terms of simplicity and applicability within the Smart Selection framework.