Follow:

Stable Isotope Analysis of Residual Pesticides via High Performance Liquid Chromatography

https://doi.org/10.3390/molecules27238587

2.2. δ13C Variations in Standard Material: Before and after HPLC Purification

To validate the reliability of isotope information by HPLC/EA–IRMS, two aspects were considered in this study: (1) the relationship of amplitude (i.e., amount of analyte transferred to tin capsule) to the carbon isotope value and (2) the effects of HPLC performance on isotope variability. As shown in Figure 3, the δ13C of CHL before HPLC was −26.18 ± 0.09‰ (average ± SD, n = 27), which was approximately 0.85‰ different than that after the HPLC fraction of -27.03 ± 0.30‰ (n = 18). Slightly lower isotope values after HPLC were also observed for CYN. That is, the carbon isotope composition of CYN after the HPLC fraction was on average 0.44% lower (average ± SD: −26.19 ± 0.65‰, n = 21) than that before the HPLC fraction (average ± SD: −25.75 ± 0.08‰, n = 24).
Figure 3. Effects of HPLC analytic procedure on determining isotope measurement in CHL standard (a) and CYN standard (b).
This isotopic difference might be related to the amplitude (or injection amount) of our target pesticides (Figure 3). The amplitude range of the CHL standard in our study was from 2.31 nA to 13.38 nA (the corresponding weights of standard material per tin capsule ranged from 0.01 mg to 0.13 mg), which was much broader than the amplitude from 1.52 nA to 2.48 nA (the estimated weight 0.03 mg to 0.08 mg) of CHL standard after the HPLC fraction. Thus, the amplitude of CHL from the HPLC fraction was close to the lowest level of analyte injection weight (mg) of powder CHL. The CYN standard results showed amplitudes per tin capsule from 4.32 nA to 14.15 nA (corresponding weight of the standard material from 0.01 mg to 0.15 mg), while the amplitudes were from 1.50 nA to 5.08 nA for CYN after the HPLC fraction (corresponding to their estimated weight from 0.03 mg to 0.15 mg per tin capsule). Moreover, SD was slightly different before and after the HPLC fraction (CHL: 0.09 vs. 0.30; CYN: 0.08 vs. 0.65). Higher precision in EA–IRMS relative to HPLC/EA–IRMS was also reported in other studies using other organic compounds, e.g., 0.08 vs. 0.16 in phenylalanine 11. Such isotope variability dependent on the sample amount (amplitude by IRMS) was widely reported in IRMS platforms. For instance, metabolite desphenylchloridazon formed by degradation of the herbicide chloridazon showed at least a 1‰ difference in δ13C values depending on the injection amount of C, while a larger standard deviation is reported with a smaller injection amount [13]. The overall result confirms that HPLC performance after solvent removal by N2-drying did not significantly change within a 1‰ difference for isotope measurements. To reduce the analytic errors from our EA–IRMS system, we also suggest a detection requirement of analyte per tin capsule (amplitude > 3 nA) in the HPLC/EA–IRMS method. Ultimately, securing the appropriate analyte amount in a tin capsule (approximately >0.03 mg) is important to produce reliable isotope compositions of specific compounds present in the environment by HPLC/EA–IRMS applications.

2.3. Application of HPLC/EA–IRMS and δ13C Determination of Target Compounds in a Soil-Crop System

The pre-treatment procedure in samples of interest for CYN and CHL detection traditionally is known to follow solid phase extraction (SPE) based on silica [24]. However, our pre-treatment procedure involved SPE extraction combined with HPLC separation (see Section 3). To evaluate the needs of the complex pre-treatment procedures, the analytic procedure was applied to a CYN-based product (11 mg/mL), a CHL-based product (9 mg/mL), and pesticide-free samples (topsoil and plant parts from a local market). Briefly, HPLC chromatograms showed that pesticides were present for CHL at 8 min and for CYN at 12 min under 45% ACN as a mobile phase, at least 2 min from the peaks of the pesticide-free sample matrix (interfering substances, present < 3.5 min) (Figure 4). This suggests that single SPE extraction (generally adopted for pesticide quantification) would experience unintended contamination from sample matrix effects when isotope values of pesticides are measured by an EA–IRMS system due to a lack of chromatographic separation ability. Consequently, SPE extraction after HPLC performance for collecting pesticide fractions would be essential to effectively exclude the interfering compound (‘matrix’ effect) and isolate residual pesticides, particularly from crop samples.
Figure 4. Chromatograms overlay in HPLC of two pesticides standards, cyantraniliprole (at 8.4 min) and chlorantraniliprole (12.3 min), and other pesticide-free samples of soil, leafy parts of lettuce, and root parts metrics (before 6 min) with constant flow (1mL/min) of 45% Acetonitrile in DW as mobile phase.
The percentage of CHL in soil decreased to 50.75 ± 0.02% (amount estimated as 0.25 ± 0.03 mg/g) and that of CYN decreased to 80.13 ± 1.22% (estimated concentration 0.80 mg/g). In contrast to the concentration-based results (Figure 5), there was no significant change in δ13C in pesticides from soils, only a slight increase in δ13C (on average 1.7‰ for CHL from the soil) at 45 days relative to the initial time (−29.90 ± 0.02‰, Table 1).
Figure 5. Estimate of amounts of residual pesticides in soil and lettuce leaves from indoor crops over a 45-day culture experiment (N = 3).

Leave a Comment