Investigation and Validation of Detection of Storage Stability of Difenoconazole Residue in Mango

Analysis & Testing Center, Chinese Academy of Tropical Agricultural Sciences, Xueyuan Road 4, Haikou 571101, Hainan, China Laboratory of Quality & Safety Risk Assessment for Tropical Products (Haikou), Ministry of Agriculture, Xueyuan Road 4, Haikou 571101, Hainan, China Hainan Provincial Key Laboratory of Quality and Safety for Tropical Fruits and Vegetables, Xueyuan Road 4, Haikou 571101, Hainan, China Quality Supervision and Inspection Center of Tropical Agro-Products, Xueyuan Road 4, Haikou 571101, Hainan, China


Introduction
Difenoconazole is a synthetic fungicide which acts by inhibiting demethylation of fungus during the synthesis of ergosterol [1,2] (the molecular structure is shown in Figure 1).It is widely used for crop protection against fungal pathogens, such as ascomycetes, basidiomycetes, and deuteromycetes [3][4][5].Although it degrades rapidly, low level of residue in agrofood can threaten human life especially in raw foods, such as vegetables and fruits [6][7][8].us, determining the pesticide residue is an e ective way to guarantee agrofood safety.Mango is one of the most popular fruits, and its farming is threatened by fungal diseases.Consequently, difenoconazole is often applied to overcome this.Codex Alimentarius Commission (CAC) regulates the MRL of difenoconazole in mango to be as low as 0.07 mg/kg [9].
Gas/liquid chromatography (GC/LC) and gas/liquid chromatography-mass spectrometry (GC/LC-MS) methods are mainly used for determination of difenoconazole [10][11][12][13].Besides being accurate, these methods are timeconsuming due to complex processes involved, and thus, a large number of samples need to be stored for analysis.During storage, although samples are stored in relatively low temperature, their components can exhibit drastic changes due to their volatility, hydrolysis, photolysis degradations, and enzymatic reactions.Nevertheless, it is not clear what happens to the pesticide residues under di erent storage conditions.
erefore, it cannot be concluded that the original level of the residue remains constant in storage, or errors do not occur during detection.However, there are few studies on stability of the pesticide in stored samples and insu cient data to explain this, hence making it a blind spot to detection.
Evidently, storage is a part of the detection course, and to analyze the stability of pesticide in samples accuracy test method is critical [14,15]. is is also necessary to guarantee the safety of agrofood.e accurate information on residual characteristics is important for the application of pesticides.To the best of our knowledge, the storage stability of difenoconazole in mango has not been reported.
In this work, a method to determine stability of difenoconazole residues in mangoes was established, and then the residue of difenoconazole in mango stored under different conditions was dynamically studied. is work investigated the storage stability of the difenoconazole in mango and the main e ects on its stability.Reasonable storage conditions for samples were proposed to improve the accuracy of detection.is guarantees the data validity of pesticide residue test and eliminates potential errors.Storage stability data also perfected the evaluation system of the pesticide characteristics.
is would be e ective guide to reasonable application of pesticides and measures to guarantee quality and safety of agrofood.

Reagents and Instruments.
Standard difenoconazole was purchased from Dr. Ehrenstorfer GmbH (Augsburg, Germany).HPLC grade acetonitrile (ACN), methyl alcohol (MeOH), ethyl acetate, and acetone were obtained from Fisher ( ermo Fisher, USA).Primary secondary amine (PSA) was supplied by Agilent Technologies (Santa Clara, CA, USA).Stock standard solution (1000 mg/L) of difenoconazole was prepared in MeOH.Series of working standard solution were prepared in stock solution and then diluted with MeOH.NaCl and MgSO 4 were obtained from Guoyao Chemicals Co., Ltd.(Shanghai, China), and puri ed water was prepared using Milli-Q apparatus (Ultima Duo 200, COMBI).

Development of Detection Method for Difenoconazole
2.2.1.Sample Pretreatment.666 m 2 eld of mango trees in Danzhou city of Hainan province were employed in this study.In a whole growing season, no difenoconazole was used on these trees.Mango fruits were uniformly collected from 12 sampling points in the eld and assigned blank samples.After homogenization, the mango sarcocarps were uniformly spiked with the standard solution at three di erent concentrations (0.01, 0.05, and 0.1 μg/g) and kept open for 30 min for the evaporation of the solvent.Half an hour later, the spiked samples were mixed.10.0 g of the spiked samples and 20 mL of the di erent solvents (acetonitrile, ethyl acetate, and acetone) were homogenized for 2 min and then centrifuged at 8000 rpm/min for 5 min.e supernatants were transferred into a centrifuge tube with 5.0 g sodium chloride and homogenized for 1 min.e mixture was centrifuged at 8000 rpm for 2 min.10 mL of the resultant supernatants was added into a mixture of 0.8 mg MgSO 4 and di erent concentrations of PSA and homogenized for 1 min [16].en, the mixture was centrifuged at 8000 rpm for 5 min, and 5 mL of the supernatants was ltered using 0.22 μm organic phase lter membrane for assays.
e gradient elution program was as follows: 0-3 min, 10% A; 3-5 min, 90% A; 5-5.2 min, 10% A; and 5.2-7 min, 10% A. Positive electrospray ionization mode (ESI + ) was run using multiple reaction monitoring (MRM) with two mass transitions where the target analytes yielded [M + H] + precursor ions.Two of the strongest daughter ions were used for quanti cation and con rmation in the assay.
e analyst software platform enabled instrument control and data processing using version 1.6.2 of AB SCIEX mass spectrometry system.

Validation of the Method.
e precision, stability, and sensitivity of the method were evaluated.Blank samples were used for method validation [17].A linear calibration curve was constructed by plotting targeted pesticide peak areas ratio against the concentration values.Linearity, representing the usability, was assessed as a coe cient (R 2 ) calculated from ve points (0.005, 0.01, 0.020, 0.030, and 0.050 μg/mL) of the calibration curves.e recovery for the precision of the method was evaluated by carrying out ve consecutive extractions (n 5) of the spiked samples at three concentration levels (0.01, 0.05, and 0.10 mg/kg).e RSD of these results re ects the stability of the methods.e LOD and LOQ were determined according to the signal-to-noise ratios 3 and 10, which indicated the sensitivity of the method.

E ect of Matrix on the Method.
e matrix in the sample would a ect the detection of the target compounds.
us, matrix e ect (ME) of the sample was also investigated to evaluate the method.Blank samples were treated as above, and standard stock solution was added into the extraction solution at three di erent concentrations (0.01, 0.05, and e analysis was conducted 5 times.

Analysis of the Uncertainty of the Method.
According to the characteristic results in this work, evaluation of uncertainty of type A showed this method was suitable for detection.Bessel formula (2) was used to evaluate the method.0.05 μg/mL of the spiked samples was evaluated using the developed method.e analysis was conducted 5 times.e uncertainty (u) was calculated by using the following formulas: (2)

Storage Temperatures.
Stored samples were spiked with the pesticide as described above and then stored at different temperatures (−20 °C and 4 °C).For samples at −20 °C, detection of residues was done at the intervals of 0, 1, 3, 6, and 12 months and 1, 2, 3, and 4 weeks for 4 °C.Quality control samples, prepared by adding standard into the blank samples at the same concentrations, were analyzed simultaneously.Detections were all repeated three times.

Storage Modes.
Typically, mango sarcocarp and analyte solutions are preserved at low temperature in actual detection operation.In this work, these modes of storage were also investigated.e samples were pretreated as before and the extract preserved at 4 °C for 2, 4, 6, and 8 weeks hermetically.e solution would be topped up if the solvent levels went down.e quality control and blank samples were also prepared in the same way.

Statistical Data
Analysis.Data were expressed as mean ± standard error (SE).Variance analysis was conducted by SPSS Statistics (version 19.0, IBM).Results of inhibition of seedling growth were analyzed by one-way ANOVA (analysis of variance) and Fisher's LSD (least significant difference) test at probability levels of 0.05 and 0.01.

Selection of the Extraction Solvent.
To choose an appropriate type of extraction solvent, three types of solvents were compared: ethyl acetate, acetone, and acetonitrile.Low recoveries, from 55% to 65%, were obtained when ethyl acetate was used (Table 1).e color of acetone extract was darker than that of acetonitrile, meaning that more impurities were obtained.High levels of impurities will affect the purification and the detection system.us, acetone was not an appropriate extraction solvent although having 90-105% recoveries (Table 1).At the same time, less impurity was extracted from the matrix by acetonitrile.Excellent recoveries of the residue extracted using acetonitrile had low RSDs, 4.0-7.0%.erefore, acetonitrile was used.
Detection of pesticide residue mainly involves sample preparation (extraction and purification) and instrumental analysis.Outstanding sample preparations should be rapid, simple, inexpensive, and environmentally friendly.For detection, extraction is the first and most crucial step, needing less impurities and highly targeting.Perfect matched polarity between solvent and analyte is important to improve the extraction efficiency and minimize the interference from the matrix.In this work, ethyl acetate was immiscible with water phase, and this inhibited the permeation between solvent and target compounds.Consequently, lower extraction efficiency was obtained.Accordingly, acetone and acetonitrile are miscible with water, and this improves the extraction efficiency.At the same time, acetonitrile extracted less impurities than acetone.ereby, acetonitrile was used as a solvent for accuracy of the column and instrument.

Sample Purification Process.
e effects of different amounts of PSA (50, 75, 100, 125, and 150 mg) on the detection method were investigated.Figure 2 shows that excellent recoveries were obtained when 75-125 mg PSA was employed and that excess or insufficient of its amounts would lead to poor recoveries.Insufficient PSA could not easily absorb the impurities in the samples to produce detectable matrix effects.Excess PSA lowered the recoveries.Generally, the optimal amount of PSA in the test was 100 mg.In the detection of contamination, solid phase extraction (SPE) column purification method involves complex operations. is is a key process which determines the efficiency as it involves series of physical reactions such as adsorption and desorption [18].Compared with the SPE method, the dispersive-solid phase method for cleanup was chosen for this work [19][20][21].It is flexible and involves simple processes, and it also would serve as a modification template depending on the analyte properties, sample composition, equipment, and analytical techniques [22].

Optimization of Instrument Condition.
Standard solutions of difenoconazole were used to optimize the conditions to develop excellent analytical method.Prior to the analysis of the samples, target fragmental ions of difenoconazole were monitored in 100-500 m/z scan range so as to obtain the best spectrum of the positively charged precursor ions.
Journal of Food Quality e results showed that 251.00 and 337.00 m/z ([M + H] + ) were the two strongest daughter ions of difenoconazole.Reversed-phase HPLC is appropriate for difenoconazole due to its polarity.Proper retention time, perfect shaped peak, high resolution, and strong intensity molecular ion peak were obtained after optimization.
To obtain high intensity, declustering potential (DP) voltage and collision energy (CE) for precursor ions were optimized (Figure 3).CE a ects ions in a more intense way than DP on optimization.Initially, intensities of ions quantitatively (251.00) and qualitatively (337.00) were highest when DP was xed at levels of 50-150 V and CEs at 32.98 V and 35.70 V.
en, CEs were xed and DP was adjusted to 125.9 V to obtain the highest ion intensities.Quantitative ion chromatogram of difenoconazole in quality control samples, blank samples, storage samples, and standard solution is shown in Figure 4.

Accuracy and Stability of the Method.
Linearity shown in this work is from 0.005 to 0.050 μg/mL with a good correlation coe cient of 0.9967.Recovery analyses were carried out for three concentrations (0.01, 0.05, and 0.1 mg/ kg).Precision and accuracy of the method were expressed by determining the recoveries and their relative standard deviations (RSD, %), as shown in Table 2. e results exhibited recovery rates of 70% to 105% with less than 9% RSDs consistent with determined residues in the samples [9,23].
ese results ascertain that residues in the extract can be accurately determined at concentrations between 0.005 and 0.05 μg/mL.

Sensitivity of the Method.
Difenoconazole standard solution was added into blank mango samples to determine the LOD and LOQ of the method.e residue concentrations were considered to be the LOD and LOQ of the equipment when the quantitative ion peak heights were 3 and 10 times to the noise heights.According to the results, LOD and LOQ of the method were 0.002 and 0.01 mg/kg, respectively.LOQ was quite lower than the MRL (0.07 mg/ kg) which is regulated by CAC, and hence this is the most competent method to determine the concentrations of the residue.

Matrix E ect on the Method
. Matrix e ects, at different concentrations, are shown in Table 3.All MEs were over 100%, manifesting that the components in the matrix promoted the ionization of the difenoconazole.e trend shows that the e ects increased with the target compound decrease.ME a ected the detection accuracy but this could be eliminated by developing a calibration curve using the blank sample extract solution.e data were obtained based on this method.Journal of Food Quality

Measurement of Uncertainty of the Method.
e recovery results in Section 3.4.1 (0.05 μg/mL) were used for calculating the uncertainty according to the formula in Section 2.2.5.e uncertainty of the method was calculated as 0.001, and its relative standard uncertainty was 0.025. is value is quite low and re ects the high accuracy and stability of this method.According to the mathematical model, the uncertainty was due to several aspects, such as the repeatability of detection, the accuracy of the standard, the accuracy of the constant volume, and the measurement of the initial samples.
e uncertainty is a comprehensive embodiment of all the processes and operations.

Storage Stability at Di erent Temperatures.
e results of storage stability of residue in mango sarcocarp at low temperature are shown in Figure 5.
ere were low    Journal of Food Quality variations of residue concentrations when the samples were stored at −20 °C.Twelve months later, only about 8.63% and 0.52% of the residue degradation of 0.1 mg/kg and 0.010 mg/ kg concentration occurred.At 4 °C, the concentration of residue remained more than 86% for one week, after which it exhibited a downward trend.In three weeks, only 66%, 43%, and 68% of the residue at the three concentrations, respectively, remained in the samples.is indicates that storage temperature is a critical factor a ecting the stability of the residue.e results show that low storage temperature greatly inhibited the degradation of difenoconazole, and the process is promoted by relatively high temperature (4 °C).
us, samples can be kept at −20 °C for a long time before analysis, and 4 °C is only appropriate for temporary storage.
Residue degradation in preserved samples was mainly a ected by chemical and biochemical factors.Sample matrix has abundant enzymes and microbes, probably causing degradation of the residue but this slows down when the samples are stored at low temperature.In contrast, relative high temperature would enhance their activities, which would promote the degradation of pesticide.Chemical degradation, such as hydrolysis, oxidation, and photolysis, was also in uenced by the temperature.e results were consistent with the reports [24,25].erefore, relative high temperature promotes the residue degradation.However, the main factor that a ects the stability of this residue remains unclear, and this calls for further investigations.

Storage Stability in Di erent Storage Modes.
To further clarify the cause of degradation of the residue, more investigations were conducted.Samples were thoroughly pretreated, and the concentrated residue was stored in the solvent. is is also a frequently used storage mode of the sample in actual analysis operation.Results indicated that there were little changes (less than 10%) of the residue concentration for 6-8 weeks at 4 °C although much of the solvent had evaporated (Figure 6).In the solvent, enzyme and microbe would lose their activity, excluding biochemical degradation, and chemical degradation would be the only  e results showed that little chemical degradation occurred to the residue, although not fully exhibited.us, biochemical factors would be determined as the main reason for the degradation of the pesticide residue.

Conclusion
Results from this work show that biochemical factors mainly cause the degradation of difenoconazole, and the factors are sensitive to temperature.It is a stable mode for the residue preservation to store the extracted samples in the solution.
ese ndings not only provide basic data on the storage of the difenoconazole in the samples but also con rm the validity of the analysis.
e investigation eliminates the potential errors in detection and also promotes a reasonable guidance for application of pesticides.All these measures are important in guaranteeing food safety and quality.

Figure 3 :
Figure 3: e optimization of the DP voltage and CE.

Figure 5 :
Figure 5: Storage stability of difenoconazole in mango stored at −20 °C and 4 °C over time.Concentrations are presented as the mean percentage compared to initial values.Values signi cantly lower than the controls are indicated with a single asterisk (LSD test, * p < 0.05) or double asterisk ( * * p < 0.01).Error bars are standard errors of the means (n 3).

Figure 6 :
Figure 6: Storage stability of the difenoconazole concentration in mango extract solution at 4 °C over time (weeks).Concentrations are presented as the mean percentage compared to the initial values.Values signi cantly lower than the initial controls are indicated with a single asterisk (LSD test, * p < 0.05) or double asterisk ( * * p < 0.01).Error bars are standard errors of the means (n 3).
Standard solution in purified reagent (MeOH) was labelled A, and spiked solution was labelled B. A and B solutions of same concentration were analyzed by the UPLC-MS/MS method.e peak areas of the quantitative ion were designated as area A and area B.

Table 1 :
Results of recoveries from di erent solvents (n 3).