Introduction

Pesticides are widely used throughout the world to protect agricultural and horticultural crops by preventing, or lessening the damage caused by pests. Despite their beneficial effects on agricultural yield, there are also significant problems for their use or misuse such as occupational exposure of farmers, accidental exposure of bystanders and potential toxicity to humans owing to accidental or intentional exposure [1]. In such cases, blood samples may be sent to a toxicological laboratory for diagnosing a poisoning.

For decades, many methodologies have already been established for the determination of pesticides, including immunoassay [2], electrochemiluminescence biosensor [3], LC–MS [4] and so on. However, each of these methods has their own shortcomings, such as time-consuming sample preparation, sophisticated instrumentation, low recovery and reproducibility, which limit their use as diagnostic tools. GC–MS is an excellent tool for qualitatively and quantitatively detecting pesticides in blood. Because blood is a complex matrix, and pesticides in blood are usually at low concentrations, the separation of target analytes and reduction or elimination of interference in blood have become the most critical steps in the entire analytical process [5]. Traditional sample preparation technologies for pesticide detection include solid-phase extraction (SPE) [6], solid phase micro extraction (SPME) [7], liquid–liquid extraction (LLE) [8], microwave extraction (MAE) [9], accelerated solvent extraction (ASE) [9], supercritical fluid extraction (SFE) [10] and so on. However, these methods have their own shortcomings, such as being time consuming and/or labor intensive, exposing workers to hazardous solvents and for some of them are difficult to easily and simultaneously achieve high quality for the vast majority of pesticide extraction and analysis [11]. Therefore, a more efficient, accurate and easier extraction method is required imminently.

QuEChERS is a type of sample preparation method that was first reported by Anastassiades et al. in 2003 [12]. This method has many advantages, including high recovery for a wide scope of pesticides with different polarity and volatility in different matrices, meeting low detection limits, use of smaller volumes of organic solvents and simple operation [13]. Therefore, QuEChERS method and its modified versions are widely used in food analysis and pesticide residues determination [14, 15].

In addition, Fe3O4 MNPs, a novel and interesting material, have attracted enormous scientific attentions in the past few years. On account of their special magnetism, adsorption and extremely large surface area, they have become functionalized magnetic absorbents and have been successfully applied in separation of pesticides in different samples [16]. Fe3O4 MNPs in combination with QuEChERS sample preparation technology have been developed. Li’s group used QuEChERS sample preparation method with Fe3O4 MNPs as the adsorbing material and GC–MS/MS method established a new method for multiple pesticides determination in vegetables and fruits. Due to the unique magnetism of Fe3O4 MNPs, pesticides can be easily separated out of sample solution by an external magnetic field. Thus, the phase separation can be conveniently conducted, which is beneficial to guarantee simplicity and rapidity of the method. However, in their research, there is no comparison on the effect of adding Fe3O4 MNPs or not. Furthermore, as we know from consulting previous literatures, there is no report about GC–MS method for pesticides detection in human blood using QuEChERS method with Fe3O4 MNPs as the adsorbing material.

This work is the first attempt to establish a GC–MS method based on QuEChERS sample preparation method with Fe3O4 MNPs to analyze multiple pesticides in a blood sample. The Fe3O4 MNPs used in this study had a regular shape, close to spherical morphology and their mean size was approximately 20–40 nm. PSA (Primary secondary amine) had an irregular shape, and its mean size is about 50 μm. Fe3O4 MNPs, C18 and PSA were fabricated to a magnetic material in a simple way based on an “aggregate warp” process [14]. By coupling with GC–MS, a rapid, simple and effective method for the analysis of pesticides in blood was established. The method exhibited satisfied recovery, selectivity, stability and reproducibility, which indicating it would be a potentially alternative tool for pesticides detection in blood samples.

Experimental

Chemicals and Reagents

Malathion, methyl isofenphos, dichlorvos, chlorpyrifos, phenthoste, p,p′-DDD, p,p′-DDE, Fe3O4 magnetic nanoparticles (20–40 nm) were purchased from Aladdin (Shanghai, China). Primary secondary amine and C18 were purchased from Agela Technologies (Tianjin, China). Anhydrous magnesium sulfate, sodium chloride was purchased from Sangon (Shanghai, China). Acetonitrile (ACN) and methanol were purchased from Siyou Scientific Company (Tianjin, China). All these chemicals were used directly without further purification.

Sample Extraction Procedure

The blood samples were kindly provided by the First Affiliated Hospital of Chongqing Medical University. Drug-free blood samples were used as negative controls. 2 mL blood sample was added into an eppendorf vial (10 mL). Appropriate volumes of pesticide standards were also added to the vial, and then shaken vigorously for 1 min to ensure that the solvent interacted well with the entire sample. The samples were extracted with 2 mL acetonitrile (ACN) by the homogeniser for 30 s. Anhydrous NaCl (100 mg) and anhydrous MgSO4 (400 mg) were added to the mixture and vortexed immediately for 1 min. After centrifugation (5000 rpm, 5 min), the supernatant was transferred to another eppendorf vial (5 mL) containing 100 mg anhydrous MgSO4, 25 mg C18, 40 mg PSA and 30 mg Fe3O4 MNPs (optimized condition). The vial was shaken in a vortex mixer for 1 min, and the supernatant was collected with the aid of an external magnet, and then evaporated off to dryness under nitrogen. Finally, the sample was dissolved in 50μL of acetonitrile and 1μL of the resulting solution was injected into the GC–MS.

GC–MS Analysis

GC–MS was performed using an Agilent 7890A/5975C (Agilent Technologies, USA) with HP-5MS column (30 m × 0.25 mm i.d., 0.25 μm film thickness) at a constant flow (1 mL/min of helium). The split/splitless injector was operated in splitless mode. The GC–MS operating parameters were as follows: inlet temperature set at 260 °C, interface temperature set at 280 °C, MS source set at 230 °C, MS Quad set at 150 °C and ionization energy was under 70-eV. The analytes were monitored in selective ion monitoring (SIM) mode. The oven temperature was initially held at 100 °C for 1.5 min, and then increased to 200 °C at a rate of 20 °C/min, subsequently, it was increased to 280 °C at a rate of 6 °C/min and held for another 3.0 min. The total run time was 24 min. The Chromatographic separation of the standard mixture of seven target compounds was shown in Fig. 1. The retention times (t R), molecular weights and target ions for GC–MS analysis of the target compounds was shown in electronic supplementary material ( Table 1).

Fig. 1
figure 1

Chromatographic separation of the standard mixture of seven target compounds (2 µg mL−1) obtained in SIM

Full size image
Table 1 Retention times (t R), molecular weights and target ions for the GC–MS analysis of the target compounds
Full size table

Results and Discussion

Effect of Extraction With and Without Fe3O4 MNPs

To testify the purification effect of the proposed QuEChERS extraction method with Fe3O4 MNPs, the blood sample with different clean-up procedures were compared. As shown in Fig. 2, the blood sample only extracted by ACN looked yellow, while extracted by ACN/PSA/C18 was paler yellow. The final blood sample extracted by ACN/PSA/C18/Fe3O4 MNPs looked transparent. It is indicated that ACN/PSA/C18 displayed a better clean-up performance than ACN to remove impurities and ACN/PSA/C18/Fe3O4 MNPs had the best purification effect. This implied that PSA/C18 can effectively absorb the impurities in blood and Fe3O4 MNPs also had obvious function as an adsorbent. Figure 3 shows chromatograms for blood extract after different clean-up procedures. Comparing Fig. 3a–c, the matrix effect was significantly decreased which illustrated that the ACN/PSA/C18/Fe3O4 MNPs clean-up procedure had a better clean-up efficiency compare with the ACN/PSA/C18 clean-up procedure; ACN/PSA/C18 clean-up procedure had obvious advantages compared with ACN extraction. It is clear that Fe3O4 MNPs, PSA and C18 exhibited a high adsorption capacity for interferences in samples.

Fig. 2
figure 2

Photography of clean-up performance by different procedures: a blood extracted by ACN without other clean-up steps; b blood extracted by ACN, PSA and C18; c blood extracted by ACN, PSA, C18 and Fe3O4 MNPs

Full size image
Fig. 3
figure 3

Chromatograms for blood extract after different clean-up procedure: a SIM chromatogram for a blank blood sample with ACN clean-up; b SIM chromatogram for a blank blood sample with ACN, PSA and C18 clean-up; c SIM chromatogram for a blank blood sample with ACN, PSA, C18 and Fe3O4 MNPs clean-up

Full size image

Optimization of the Amount of Adsorbents

To achieve satisfactory purification effect, two types of absorbents, PSA and Fe3O4 MNPs were investigated for the seven pesticides (malathion, methyl isofenphos, dichlorvos, chlorpyrifos, phenthoste, p,p′-DDD, p,p′-DDE) from blood samples. The recoveries of the 7 pesticides in six different amounts of PSA (10–60 mg) are shown in Fig. 4a. With the changing PSA amounts, the maximum recoveries of these pesticides were obtained at 40 mg. When more than 40 mg PSA was added, the targets recoveries began to decrease. Thus, 40 mg PSA was selected for clean-up procedure. The recoveries of target pesticides in Fe3O4 MNPs (10–60 mg) are shown in Fig. 4b. When the amount of Fe3O4 MNPs added close to 30 mg, the recoveries reached to maximum, indicating that 30 mg was the proper amount to achieve optimal recoveries.

Fig. 4
figure 4

Effect of a mass of PSA, b mass of Fe3O4 MNPs on the extraction efficiency of the studied pesticides (2 µg mL−1 each)

Full size image

Method Validation

To verify the accuracy and precision of the method, several basic analytical parameters were evaluated. Calibration curves were calculated with a matrix-matched standard calibration in blank samples. The linear correlation coefficients (r 2) were higher than 0.9949 in the range from 0.3 to 4.0 µg mL−1. The LOD and LOQ for each pesticide were determined as the lowest concentration yielding signal-to-noise ratios of at least 3:1 and 10:1, respectively. The LODs and LOQs for the 7 pesticides were found to be between 0.011–0.163 and 0.036–0.538 µg mL−1, respectively. This method exhibits high sensitivity as well as simply and rapid properties. This highly sensitive detection of pesticides is due to the excellent clean-up performance displayed by ACN/PSA/C18/Fe3O4 MNPs. The r 2, LODs and LOQs are presented in electronic supplementary material (Table 2).

Table 2 Determination coefficient (r 2), LODs (µg mL−1) and LOQ (µg mL−1) for the seven pesticides
Full size table

Recovery dates were obtained for all pesticides spiked into blank blood at concentrations of 1.0, 2.0, 4.0 µg mL−1 and each concentration was tested in three replicates. The relative standard deviations (RSD) were also calculated. The average recoveries of the target pesticides ranged from 70.0 to 111.8%, and the RSD was not more than 10.5%. The recoveries and RSD were in line with pesticide analysis requirements. The results are summarized in electronic supplementary material (Table 3).

Table 3 Recoveries and precisions of target pesticides in blood sample at three concentrations
Full size table

Method Application

The proposed method for detection of 7 pesticides in blood was successfully applied to the investigation of forensic toxicological cases of poisoning. A 54-year-old man was found dead at his house, and a bottle with pesticide was also found close to him. His stomach content and blood were sent to toxicological laboratory for diagnosing a poisoning. Analysis of his stomach content and blood showed the presence of dichlorvos, while the analysis of blood sample revealed the presence of dichlorvos at a concentration of 3.6 µg mL−1. Compared with liquid–liquid extraction, the matrix effect was found to be decreased significantly, and the detection limit was reduced 11.2 times. Once again, it is proved that QuEChERS extraction with Fe3O4 MNPs improve the sensitivity of the method.

Conclusion

In this work, a method to identify and quantify multiple pesticides in human blood by QuEChERS extraction with Fe3O4 MNPs and GC–MS analysis is presented. The method avoids some time-consuming centrifugation or filtration, becomes simpler, more efficient and faster by adding Fe3O4 MNPs to the purification process. Based on the results presented above, the modified methodology also exhibits high sensitivity, satisfactory precision and reproducibility. Under the optimum conditions, the detection limits of the proposed method for seven pesticides ranged from 0.011 to 0.163 µg mL−1. Good linearity (R value ≥0.9949) was achieved at concentration levels of 0.3–4.0 µg mL−1, and method recoveries were found in the range of 70.0–111.8% at different concentrations. It absolutely meets the requirements for multiple pesticide determination in blood and will have broad applications in toxicology analysis.