Optimization of Glyphosate Removal from Aqueous Solution Using Shea Nut Shell Bioadsorbent; A Response Surface Methodology Approach

Glyphosate contamination of water resources from agricultural activities presents a significant environmental challenge, particularly in Nigeria where the herbicide is widely used. This study evaluated a shea nut shell (Vitellaria paradoxa)-derived bioadsorbent for glyphosate removal from aqueous solution and assessed the principal adsorption variables using response surface methodology based on a Box-Behnken design. The bioadsorbent was prepared by alkaline (NaOH) treatment followed by carbonization at 400 °C for 2 h and characterized using FTIR, SEM-EDS, PXRD, and XRF. Characterization revealed lignocellulosic functional groups (C-H, C=C, C-O-C, C-O), a rough and porous surface morphology, semi-crystalline cellulose with a crystallite size of 15.69 nm, and inorganic constituents including potassium, calcium, and iron. The point of zero charge (PZC) was 6.8. Batch adsorption experiments were conducted using a 46-run Box-Behnken design investigating pH (2–10), contact time (10–180 min), adsorbent dosage (0.10–2.00 g/100 mL), initial glyphosate concentration (10–100 mg/L), and temperature (25–45 °C). Experimental glyphosate removal ranged from 0% to 100%, with complete removal observed in two runs, the highest experimental adsorption capacity was 62.14 mg/g. The quadratic response surface model explained 52.92% of variation (R² = 0.5292) but exhibited limited predictive capability (adjusted R² = 0.1526; predicted R² = −0.8186), and the overall model was not statistically significant (F = 1.405, p = 0.2085). Initial glyphosate concentration was the only statistically significant linear factor (F = 6.106, p = 0.0206), with a significant quadratic effect (F = 7.187, p = 0.0128). The contact time × adsorbent dosage interaction approached significance (p = 0.0544).