the four matrices ranged from 81.59 and 104.92% at the low con-centration (10 ng/mL) and from 82.06 and 105.41% at the highconcentration (100 ng/mL). The RSDs were between 0.78 and 8.77%at the low concentration (10 ng/mL) and between 1.03 and 5.36%at the high concentration (100 ng/mL).Calibration standards consisted of 10 g of the food samplesspiked with the test range of 0.5–1000 ng/mL. Tables 4 and 5show the equations of the calibration curves and linear regressioncoefficients for the PAs. The calibration curves were generated bya linear least squares regression analysis of the PA compounds/ISpeak area ratio versus the amount of spiked PAs. The correlationcoefficients (R2) for all of the PA compounds were greater than0.995, indicating excellent linearity. The limits of detection (LOD)and quantitation (LOQ) were based on signal-to-noise ratios of 3(S/N ≥ 3) and 10 (S/N ≥ 10), respectively. The LODs of this methodranged from 0.07 to 0.59 ng/mL for soybeans, 0.06 to 0.43 ng/mLfor seed soil, 0.07 to 0.53 ng/mL for milk, and 0.07 to 0.60 ng/mLfor margarine. The LOQs ranged from 0.25 to 1.97 ng/mL for soy-beans, 0.20 to 1.43 ng/mL for seed oil, 0.23 to 1.76 ng/mL for milk,and 0.25 to 1.99 ng/mL for margarine, indicating high sensitivity forthe determination of trace PAs in high lipid foodstuffs.The precision and accuracy were also evaluated at three con-centration levels (10, 50, and 100 ng/mL). The results of the intra-and inter-day accuracies and relative standard deviation (RSDs) arepresented in Table S1 and S2. The accuracies of all of the analytes