Kelp (Laminaria japonica) is a popular and nutritious sea vegetable, but it has a strong biosorption capacity for heavy metals. The high content of cadmium (Cd) and lead (Pb) is a threat to the quality of kelp. The objective of this study was to investigate the effects of high-pressure-assisted extraction (HPAE) conditions on Cd and Pb removal efficiency from kelp. Pressure intensity (0.1–200 MPa), the number of HPAE cycles (one to five) and acetic acid concentration (0–10%) were optimized using response surface methodology. The pressure intensity had the most significant positive effects on Cd and Pb removal efficiency, while the correlation between acetic acid concentration and removal efficiency was positive for Cd and negative for Pb. The optimum conditions for the removal of Cd and Pb were attained at 188 MPa, with four cycles and with an acetic acid concentration of 0%. At optimum conditions, the experimental values of removal efficiency were 61.14% (Cd) and 70.97% (Pb), and this was consistent with the predicted value, confirming the validity of the predictive model.

As an ingredient in various foods, Chrysanthemum morifolium flower is popular due to its multiple health benefits. Pyrrolizidine alkaloids (PAs) are hepatotoxic secondary metabolites in Chrysanthemum family. Effects of high-pressure extraction (HPE) on PAs removal efficiency, as well as the retention efficiency of functional components, including chlorogenic acid, luteolin-7-β-D-glucopyranoside, 3,5-dicaffeyl quinic acid and total flavonoids, were investigated and optimized using response surface methodology (RSM). Pressure (0.1–200 MPa), numbers of cycles (1–5) and acetic acid concentration (0–10%) were chosen as the independent variables. The results indicated that the pressure was the most significant factors affecting all responses. The optimum HPE for removing Pas and retaining functional components were set at 124 MPa, with one cycle and with an acetic acid concentration of 10%. After comparing the experimental optimum values and predicted optimum values, the validity of RSM model was proved.