*4.3. Statistical Analysis between Physicochemical Parameters and Removal Efficiency of Ten Pesticides*

As mentioned earlier, correlation analysis (Table 5) revealed that the dominant physicochemical parameter affecting the rate of pesticide residue removal during heat treatment (blanching and boiling) was log *P* (negative correlation). As depicted in Table 1, the log *P* values of chlorfenapyr, diniconazole, indoxacarb, fludioxonil, pyraclostrobin, and lufenuron ranged between 3.99 and 5.12, and the log *P* values of azoxystrobin, chlorantraniliprole, imidacloprid, and thiamethoxam ranged between −0.13 and 2.86. Therefore, the pesticide residues for chlorfenapyr, diniconazole, indoxacarb, fludioxonil, pyraclostrobin, and lufenuron, which have relatively high log *P* values, could not be easily removed. Similarly, Nagayama reported that the log *P* value and reduction were inversely proportional based

on an analysis of the residual pesticide amount after blanching or making jam using agricultural products collected from the market [39]. Reichman et al. reported volatility changes owing to phase distribution according to Henry's constant when a pesticide is placed in a wet medium. Therefore, in their soil test, the highest volatilization rate of trifluralin was reported to correspond to the highest Henry's constant [40]. Similarly, in a study by Kwon, the reduction in pesticide residues by boiling water was proportional to Henry's constant of pesticides [27]. However, by comparing the pesticide residue reduction and Henry's constant, a proportional relationship with Henry's constant was not found. The dominant variables in the PCA plot were water solubility and Log *P* (Figure 3a–i3). These two variables are the important variables that have the greatest influence on PC1.
