*3.4. Application of Dual Isotope Surrogates to Aquatic Products*

The proposed method was applied in the analysis of a batch of different aquatic products collected from commercial markets. A total of 47 out of 136 samples were detected as positive. Four samples were found to exceed the linear range of the calibration curve with ENR-*d5* as the surrogate. Three positive samples were then selected for validation of the method; these were a common carp (*Cyprinus carpio*), a bullfrog (*Lithobates catesbeiana (Shaw)*) and a bluntnose black bream (*Megalobrama amblycephala*), which showed a medium, high and significantly high concentration of ENR, respectively.

The ENR content in three aquatic products, determined using dual isotope surrogates, is summarized in Table 4. HPLC–MS/MS analyses show that the new isotope surrogate ENR-*d3* was successfully applied and detected in positive aquatic products. As a result, the detection of ENR was achieved using the two isotope surrogates, respectively. A slight difference in the detected concentrations and the RSD values between the two quantitative results was found. As shown in Table 4, ENR contents of 108 ± 7.25, 681 ± 35.7, and 3903 ± 433 μg/kg, using ENR-*d5* for quantitation, were detected in the bluntnose black bream, the common carp (Figure 5d), and the bullfrog, respectively. Meanwhile, the RSD values were found to be 6.72, 5.24, and 11.1%, respectively. It should be noted that the SD and RSD values of the common carp and bullfrog were relatively high, suggesting that the higher ENR concentrations were comparable with the added ENR-*d5*. This effect results in relatively low repeatability and reproducibility of the measurement. On the other hand, all samples were measured with ENR at levels of 99.1 ± 0.173, 624 ± 4.95, and 4340 ± 21.2 μg/kg using ENR-*d3*, respectively, and RSD values ranged from 0.175 to 0.794%. The response of ENR-*d3* in solution 2 was comparable with ENR-*d5* in solution 1 (Figure 5e,f). Clearly, the values of the two quantitative methods agreed well with each other. Meanwhile, the calculated SD and RSD values were much lower using the dual isotope surrogate method than those quantified by ENR-*d5* alone. It is worth noting that the ENR level in bluntnose black bream shows a discrepancy of around 8% between the result with ENR-*d5* and ENR-*d3*. We assume that both results are suitable for quantitation with both ENR-*d5* and ENR-*d3*. However, higher SD and RSD values were observed for ENR-*d5* as the isotope surrogate. This result may be interpreted by the variation in isotope surrogate loss through the preparation of the sample due to lower amounts of ENR-*d5*. Therefore, the addition of more ENR-*d3* could improve the stability of quantitation due to the stable synchronous compensation effect for the analyte and isotope surrogate loss during sample preparation. On the other hand, the results of ENR-*d5* are not acceptable for formal reports, as they was obtained using a calculation curve where the linear range does not cover this value. The advantage of using dual deuterated isomers was not obvious for

the practical sample with only a low ENR residue. It does work and saves time for samples with a high ENR residue, especially when the result with single surrogates exceeds the linear range of the calibration curve. These observations suggest that the accuracy and precision of quantitative results can be achieved by choosing suitable isotope surrogates. Therefore, the experimental results demonstrate that using different levels of dual isotope surrogates could provide accurate and reproducible results in one preparation.

**Table 4.** Determination of ENR in three aquatic products using ENR-*d5* and ENR-*d3* as the isotope standards (n = 9).

