*3.5. Sample Preparation*

In the establishment of the ELISA method, sample pretreatment is very important and is directly determined the accuracy of detection results. In recent years, the existing ELISA methods for PAHs detection have mainly concentrated on environmental samples, such as water (lake water, tap water, drinking water) and air [15,32,33]. However, there is no report on the detection of PAHs in aquatic products by ELISA. As a large country of aquatic products, it is imperative for China to detect PAHs in aquatic products. The ELISA method sensitivity is according to the specific binding ability of antigen and mAb, however, the

sensitivity is susceptible to interference by fat and protein in the sample. Therefore, the samples usually need to be pretreated (extracted, purified, concentrated or diluted) before analysis to reduce matrix interference. To improve the precision of detection, according to the chemical properties, the extraction method of PYR and BaP in aquatic products was explored.

First, the matrix effect was evaluated, and blank samples were directly pretreated with PBS as previously reported [33]. The results showed that the matrix effect was very severe, even false-positive. When organic solvent extraction was performed during sample pretreatment, the matrix effect problem was solved. In this study, several extraction buffers, including acetonitrile, ethyl acetate and ethyl acetate-acetonitrile (4:1; v:v), were used to investigate the extraction effect. It can be seen that a higher recovery rate was obtained by ethyl acetate-acetonitrile (4:1; v:v). Acetonitrile could not extract the drug, and the recovery of ethyl acetate was less than 70%. Therefore, ethyl acetate-acetonitrile (4:1; v:v) was selected as the extraction solvent for sample pretreatment.

### *3.6. IC-ELISA Performance for Aquatic Products*

Based on the optimization of ic-ELISA conditions and sample pretreatment described above, 20 negative samples were collected for ELISA performance measurement. The LODs of the method for PYR and BaP in fish, shrimp and crab ranged from 0.43 to 0.54 μg/kg and 0.92 to 0.98 μg/kg, respectively. Meanwhile, the recoveries of PYR and BaP in fish, shrimp and crab were 81.5–101.9% and 84.9–94.0%, respectively, and the CV was less than 11.7% (Table 2). The minimum detection levels of PYR and BaP in aquatic products were far below the MRLs.


**Table 2.** The LODs, LOQs, CV and recoveries of PYR and BaP in aquatic products.

To validate the reliability of this experimental method, the ic-ELISA and HPLC-FLD methods were used to detect positive fish samples contaminated with PYR. The liquid chromatogram of PYR in the fish sample is shown in Figure S3. The results from the ic-ELISA were consistent with the results of HPLC-FLD, and the correlation coefficient was R2 = 0.9961 (Figure 5). The results indicated that ic-ELISA can be used as an available tool for the detection of PYR and BaP in aquatic products samples.

**Figure 5.** Correlation between ic-ELISA and HPLC-FLD for PYR determination of positive fish sample.
