OPTIMIZATION AND APPLICATION OF ATMOSPHERIC COLD PLASMA IN CASHEW NUT PROTEIN CONCENTRATES

The conventional isoelectric precipitation (IEP) method for protein extraction is limited by long processing times, requiring two resting steps totaling 180 minutes per batch (TR₁ = 120 min; TR₂ = 60 min). This study sought to optimize cashew nut kernel processing using response surface methodology (CCRD – Central Composite Rotatable Design, 13 experiments; 2²). The independent variables were TR₁ (0–120 min) and TR₂ (0–60 min.), and the responses were yield (%) and protein content (%). The objective was to reduce extraction time without compromising performance. We also evaluated post-processing strategies to improve the techno-functional attributes of the optimized protein concentrate by applying cold plasma treatments: dielectric barrier discharge – DBD (50, 500, and 1000 Hz) and vacuum plasma (10 and 30 min). Functional metrics included solubility at pH 7.0, emulsifying capacity, emulsion stability, water and oil absorption, and foaming. In all runs of the design, the yield ranged from ≈47–57% and the protein content from ≈67–72%, with R² values of 0.54 and 0.91, respectively. For protein concentration, F_{regression/residuals} was ≥F_tabulated and F_{lake-of-fit/pure-error} was ≤F_tabulated, supporting the adequacy of the model. To validate the experimental model, experiments were performed at the optimum point (TR₁ = 0 min; TR₂ = 10 min), which produced ≈69±0.3% protein and ≈47±0.05% yield. The amino acid profile of the optimized concentrate corresponded to that of the standard process (TR₁ = 120 min; TR₂ = 60 min), indicating that the nutritional quality of the protein was preserved despite the drastic reduction in retention times. The results revealed that vacuum plasma, especially at 10 minutes, maximized emulsion stability (≈114 min.) and solubility (≈57%). These effects are consistent with partial unfolding and increased exposure of hydrophilic groups, which favor protein-water interactions and interfacial stabilization. In contrast, higher-frequency DBD reduced solubility and emulsifying properties, suggesting aggregation driven by excessive oxidation and the formation of less functional protein-protein networks. It is concluded that the optimized route reduced the total processing time from 180 to 10 minutes, maintaining the protein content and amino acid profile quality of the standard method. When combined with vacuum plasma post-treatment, the process produces more soluble and stable concentrates suitable for beverage and emulsion formulations. From a processing standpoint, the reduced dwell time implies lower energy demand, facilitating scale-up. The results indicate technological feasibility and industrial potential.