3.1.1. Assay Optimization

For the development of the competitive indirect ELISA for determination of carbendazim, a benzimidazole conjugate was used as a solid-phase reagen<sup>t</sup> in combination with a rabbit polyclonal anticarbendazim antibody, both developed in-house as previously described [17]. The optimum concentration of the conjugate for coating was determined by preliminary titration experiments using mixtures of the anti-carbendazim polyclonal antibody with carbendazim standards prepared in assay buffer. As shown in Figure S1a, zero standard signal values in the range 1.0–1.5 (which are considered optimum for an ELISA) were received for the following combinations of conjugate/antibody concentrations 0.5/4.0, 1.0/2.0, and 2.5/1.0 in μg/mL. These combinations were tested further regarding not only the zero standard signal but also the signal received in presence of 200 ng/mL carbendazim. From Figure S1b, where the percent absorbance values are presented, it can be concluded that the combination that provided the highest percent signal drop in presence of carbendazim was 1.0 μg/mL benzimidazole conjugate and 2.0 μg/mL anticarbendazim antibody. Thus, this combination was adopted in the final ELISA protocol. Different assay buffers were then tested including a 10 mM PBS buffer, pH 6.5, 10 mM PBS, pH 7.4, 50 mM Tris-HCl buffer, pH 7.8, and 50 mM Tris-HCl buffer, pH 8.25, and the results obtained are shown in Scheme S1 (see Supplementary Material). All buffers contained 0.4% BSA. It was found that 10 mM PBS, pH 7.4, containing 0.4% (*w/v*) BSA, was the optimum buffer for the immunoreaction between the primary antibody and the antigen since it provided the highest signals for zero standard and detection sensitivity compared to the other buffers tested. The implementation of a preincubation step of the antibody with the standards prior to addition onto the biofunctionalized wells was investigated as a means to improve assay detection sensitivity. As shown in Figure S3, the assay sensitivity, expressed by the slope of the linear segmen<sup>t</sup> covering the two standards of the lowest concentration in the calibration plot, ranged from −0.19 [dS/(ng/mL)] for 0 min, −0.26 [dS/(ng/mL)] for 30 min, −0.42 [dS/(ng/mL)] for 60 min, and −0.38 [dS/(ng/mL)] for 120 min. These results indicate considerably improved assay sensitivity when a preincubation up to 60 min was used. However, longer preincubation did not result in any additional improvement. Thus, 60-min pre-incubation was adopted in the final protocol. Under optimal conditions, the detection limit of the assay (LoD, calculated as the carbendazim concentration corresponding to percent signal value equal to 100-3SD of 16 measurements of zero standard) was 20 ng/mL. The quantitation limit (LoQ) was calculated as the carbendazim concentration for which the mean absorbance value + 3SD is equal or lower than the mean zero standard value—3SD (LoD), and was 50 ng/mL. The dynamic range of the assay was from 50 ng/mL to 2 μg/mL.

The ELISA assay presented here differs from that previously described [17] in the configuration followed. More specifically, a biotinylated secondary antibody along with streptavidin-HRP have been used, instead of an HRP-labelled secondary antibody, while TMB (instead of 2,2-azino-bis(3-ethylbenzothiazoline-6-sulfonic acid) diammonium salt— ABTS) has been employed as a chromogen (Scheme S1). Moreover, a preincubation step was adopted so as to further increase the detection sensitivity.

#### 3.1.2. Evaluation of the Anti-Carbendazim Antibody Specificity

The cross-reactivity of the anti-carbendazim antibody with four commonly reported pesticides in fruit juices [20–22], including carbaryl, imazalil, atrazine, and paraquat was tested. In Figure 2, the calibration plots obtained for each one of the four pesticides are provided. As a rule, compounds that do not provide at least 50% decrease in signal under the above described conditions, are not considered as cross-reacting materials. Thus, as shown, no cross-reactivity with any pesticide tested could be detected.

**Figure 2.** Calibration plots obtained with carbendazim (black squares), carbaryl (red circles), imazalil (blue up triangle), atrazine (green down triangle), and paraquat (purple rhombus) standards with concentrations ranging from 0 to 5000 ng/mL. All standards were preincubated for 60 min with the anti-carbendazim antibody prior to addition in the microwells. Each point is the mean value of 3 replicates ± SD.
