**3. Materials and methods**

#### **3.1 Chemicals**

Analytical IMI at purity 99.9% (from Riedel-de Haën, Germany) and 6-CNA (chemical purity 99%, from Acros-Organics, USA) were used in this tudy. IMI's chemical structure and some selected physicochemical properties are shown in Figure 4 (Tomlin, 2001). Stock standard solutions of IMI and 6-CNA (1 mg/mL) were prepared by disolving the required amount in HPLC grade acetonitrile and stored at 4C. All other chemicals used were analytical grade, except acetonitrile which was of HPLC-grade (J.T.Baker, Holland). Sodium pyrophosphate, sulphuric acid, potassium dichromate, sodium hydroxide, sodium acetate and calcium chloride were purchased from Kemika (Croatia), while ammonium acetate, mercury chloride and methanol were from Alkaloid (Macedonia).


#### **3.2 Experimental soils**

Four agricultural soil samples, having different characteristics, from two coastal regions of Croatia, namely Istria and Kvarner, were used in this study. All soils were collected from the A horizon at depths of 0-30 cm following the standard methodology of soil sampling (USEPA, 2000), air-dried for 24 hours, ground (porcelain mortar + rubber pestle) and passed through a 2-mm sieve prior to use. They were selected on the basis of their texture (mechanical composition), pH values, OC content and CEC. The soils have never been treated with IMI, as verified by analyzing its residues in the soil. Selected physicochemical properties of the tested soils are given in Table 1.


*<sup>a</sup>* cation exchange capacity; *<sup>b</sup>* hydrolitic acidity; *<sup>c</sup>* organic carbon content.

Table 1. Physicochemical properties of the tested soils.

496 Pesticides in the Modern World - Risks and Benefits

The main metabolites of IMI which have been identified in the soil include IMI-urea, 6- CNA, and 6-hydroxynicotinic acid (Rouchaud et al., 1996), which ultimately degrades to CO2 (Scholz & Spiteller, 1992). For instance, depending on the soil type, IMI labeled with imidazolidin-14C had a maximum mineralization to CO2 of 8.8% or 14% after incubation for 12 weeks (Anderson, 1995; as reviewed in Mulye, 1996). In soils, when conditions were anaerobic and without light exposure, IMI was found to be readily decomposed, resulting in desnitro-IMI as the main transformation metabolite (Heim et al., 1996; as reviewed in Mulye, 1996). The desnitro-IMI produced under dark, anaerobic conditions has been found to be more persistent than its parent compound (Fritz & Hellpointner, 1991; as reviewed in Mulye, 1995). The major transformation products resulting from incubation under nonsterile, aerobic conditions and light exposure were desnitro-IMI, IMI-urea, 6-CNA and an unknown compound. Both desnitro-IMI and IMI-urea are highly water soluble, with solubility of 180 – 230 g/L and 9.3 g/L at 20°C, respectively (Krohn, 1996a, 1996b; as reviewed in Mulye, 1996), which is much higher than IMI's solubility, while 6-CNA has been

Analytical IMI at purity 99.9% (from Riedel-de Haën, Germany) and 6-CNA (chemical purity 99%, from Acros-Organics, USA) were used in this tudy. IMI's chemical structure and some selected physicochemical properties are shown in Figure 4 (Tomlin, 2001). Stock standard solutions of IMI and 6-CNA (1 mg/mL) were prepared by disolving the required amount in HPLC grade acetonitrile and stored at 4C. All other chemicals used were analytical grade, except acetonitrile which was of HPLC-grade (J.T.Baker, Holland). Sodium pyrophosphate, sulphuric acid, potassium dichromate, sodium hydroxide, sodium acetate and calcium chloride were purchased from Kemika (Croatia), while ammonium acetate,

> N N N H N+ - <sup>O</sup> <sup>O</sup>

found to be more toxic to honey bees than IMI itself.

mercury chloride and methanol were from Alkaloid (Macedonia).

Fig. 4. Chemical structure and physicochemical properties of IMI.

Structural formula Cl <sup>N</sup>

Molecular formula C9H10ClN5O2 Molecular weight (g/mol) 255.7 Water solubility, 20° C (g/L) 0.51 Log KOW 0.57 Sorption coefficient, *KD* (L/kg) 2.46 *pKa* 11.2

Four agricultural soil samples, having different characteristics, from two coastal regions of Croatia, namely Istria and Kvarner, were used in this study. All soils were collected from the A horizon at depths of 0-30 cm following the standard methodology of soil sampling (USEPA, 2000), air-dried for 24 hours, ground (porcelain mortar + rubber pestle) and passed

**3. Materials and methods** 

**3.2 Experimental soils** 

**3.1 Chemicals** 

The mechanical composition of the soil samples was determined by sedimentation using the "pipet method" (Kroetsch & Wang, 2007). Soil samples pH values were measured in a soil + deionised water and in a soil + 0.01 M calcium chloride suspension (1:2.5, w/v). The MP 220 laboratory pH meter (Metler Toledo, Germany) was used for pH determination in aqueous phase. Hydrolitic acidity (HA) was determined by the Kappen method (Hendershot et al., 2007 ), CEC was measured using ammonium replacement (Sumner & Miller, 1996), while Na, K, Mg and Ca were analyzed by Atomic Absorption Spectrophotometer (Perkin Elmer Analyst, USA). The OC content of the soils was determined spectrophotometrically (Cary 100 Bio WINUV, Varian, Australia) by dichromate method (Darrel & Nelson, 1996 ).

#### **3.3 Batch sorption-desorption experiments**

In the present study, the IMI sorption by soils was quantified using the standard batch equilibrium method (OECD, 2000). The predetermined mass of each soil (5 g), in triplicate, was equilibrated with 25 mL of aqueous solutions of IMI by shaking in an rotary agitator (Unimax 1010, Heidolph, Germany) at 20 (±1)° C for 48 h to achieve equilibrium. The equilibrium time was determined according to previous sorption kinetics studies of the IMI sorption (Capri et al., 2001; Nemeth-Konda et al., 2002). Initial insecticide solutions, in the concentration range of 0.1, 0.25, 0.5, 1, 2.5, 5, and 10 mg/L respectively, were prepared in the background 0.01 M calcium chloride and 100 mg/L mercury chloride solution from stock IMI solutions prepared in HPLC-grade acetonitrile. Calcium chloride solution was used as background electrolyte in order to minimize ionic strength changes and to promote flocculation. Mercury chloride was added to the pesticide solution as a biocide to prevent any microbial activity during the sorption experiment. After equilibration, the suspensions were centrifuged at 4000 rpm for 30 min at 20 (±1) °C (BR4i Multifunction, Thermo electron corporation, France) to separate the liquid and solid phases. After filtration through a polypropylene hydrophilic filter of 0.45 µm (Whatman, Puradisc 25 TF, USA) the aqueous phase was analyzed by High Performance Liquid Chromatography (HPLC) using a Thermo Separation Products (Spectra System, USA) liquid chromatographic system, as described in the section 3.6. Blank samples without soil were also prepared in the same way and used to account for possible losses due to the volatilization and sorption of IMI to the cuvette walls. The average system losses were shown to be consistently lower than 3.4% of the initial

Behavior and Fate of Imidacloprid in Croatian Olive Orchard Soils Under Laboratory Conditions 499

water as necessary. Three parallel soil samples (25 g) at both concentration level including unspiked controls were used for analysis of IMI residues and its metabolite (6-CNA) at

IMI and 6-CNA were extracted from soil samples according to the method of Baskaran et al. (1997). At each sampling time, a 25 g sample of spiked and homogenized soil was extracted with 40 mL of acetonitrile-water (80:20, v/v) and shaken vigorously for 2 h using a rotary agitator (Unimax 1010, Heidolph, Germany) at 20 (±1)°C. After this time every sample was centrifuged for 20 min at 6000 rpm (BR4i Multifunction, Thermo electron corporation, France) and filtered through a polypropylene hydrophilic filter of 0.45 µm (Whatman, Puradisc 25 TF, USA). The operation of shaking and filtration was repeated three times and supernatants from each extraction were pooled. The solution was evaporated to dryness on a rotary evaporator (Laborota 4002/03 Control, Heidolph, Germany). The residue was dissolved in 1 mL of mobile phase (acetonitrile-water, 20:80, v/v). Three replicates of both level, including unspiked controls, were extracted and

The concentration of IMI and 6-CNA in aqueous solutions was determined using a reversephase HPLC system (Thermo Separation Products, Spectra System, USA) equipped with a UV/VIS detector. All analyzes were performed on a Supelco reverse phase C18 column (150 mm length, 46 mm ID, 5 μm particle size). The mobile phase of acetonitrile and water (20:80 v/v) was used under isocratic conditions at a flow rate of 1.2 mL/min. The analytes were analyzed at 270 nm wavelength. The injection volume and the column temperature were 20 μL and 25 C, respectively. Under these conditions the retention times of IMI and 6-CNA

Calibration curves for both of chemicals were linear from 0.05 to 10 mg/L with regression coefficients of *R2* > 0.999 (six calibration points, in triplicate). The detection limits of IMI and 6-CNA were 0.001 mg/L and 0.003 mg/L, while the lower limits of quantification (LOQ) were 0.005 mg/L and 0.01 mg/L. The mean recoveries for IMI and 6-CNA were 91.4% and

Relationship between soil properties and sorption as well as degradation behavior of IMI was tested by a nonparametric correlation test, Kendall-Tau. Except nonparametric tests, multiple linear regression analysis was used, which combines the relationship between different soil parameters and the sorption coefficients, as well as DT50, allowing the assumption of linear models for these parameters (Boivin et al*.*, 2005). Differences in the soil sorption capacity among and within regions were analyzed using Mann-Whitney U test, while the DT50 values were tested by one-way ANOVA test with *post hoc* comparison (Tukey HSD test) to determine the effect of initial concentration and soil on the DT50 of IMI. Data are reported as mean ± standard deviations. The results were considered statistically significant at *p* < 0.05. The data were analyzed using Statistica® software package Version 7.0 and

intervals of 0, 7, 15, 30, 45, 60, 75, 90, 105, 120, 150 and 180 days after application.

**3.5 Extraction of IMI and 6-CNA from soil samples** 

analyzed by HPLC.

**3.6 Analysis of IMI and 6-CNA by HPLC** 

were 4.3 and 1.6 min, respectively.

**3.7 Statistical analysis** 

87.8 % with a relative standard deviation lover than 5%.

Wolfram Research Mathematica® software package Version 7.0.

solute concentrations, therefore no correction was required. Control samples, containing no IMI, only soil and 0.1 M calcium chloride, were used for each series of experiment. The amount of IMI sorbed to soil after equilibration was calculated from the difference between the initial and equilibrium solution concentration using the mass-balance equation:

$$
\eta\_e^{\text{sor}} = (\gamma\_i \cdot \gamma\_e) \frac{V}{m} \tag{9}
$$

where *qe sor* is the amount of IMI sorbed at equilibrium (mg/kg), *m* is the mass of soil (g), *γ<sup>i</sup>* is the initial concentration of IMI (mg/L), *γ<sup>e</sup>* is the equilibrium concentration of IMI (mg/L), *V*  is the volume of the solution (L) from which sorption occurs.

Desorption experiments were conducted on triplicate soil samples immediately after the sorption experiments with the same initial concentrations of IMI. After completing the sorption process, the supernatant (25 mL) were removed and replaced with the same volume of 0.01 M calcium chloride and 100 mg/L mercury chloride solution. After shaking for 24 h, the suspensions were centrifuged under the conditions described previously, and the concentration of IMI was determined in the supernatants using the HPLC. This desorption procedure was repeated five times for each soil sample. The amount of pesticide remaining sorbed by the soil was calculated as the difference between the equilibrium sorbed and the desorbed amount by the following equation:

$$
\eta\_e \stackrel{\text{des}}{=} (\chi\_e^{\text{sor}} \text{ - } \chi\_e^{\text{des}}) \frac{V}{m} \tag{10}
$$

where *qe des* is the amount of IMI remaining sorbed by the soil (mg/kg), *γ<sup>e</sup> sor* is the equilibrium sorption concentration of IMI (mg/L) and *γ<sup>e</sup> des* is the equilibrium desorption concentration of IMI (mg/L). The percentage of IMI desorbed was calculated as follows:

$$P^{\rm des} = \frac{\sum\_{q=1}^{5} q\_e^{\rm des}}{q\_e^{\rm sur}} \times 100\tag{11}$$

#### **3.4 Degradation experiments**

The persistence of IMI in the tested soils was studied at two concentration levels, 0.5 and 5 mg/kg under laboratory conditions at room temperature (20±1C). For fortification of the soil at 50 mg/kg level, 100 g weighted, air-dried and sieved soil was taken in a baker and 10 mL of standard solution of IMI (1 mg/mL, in acetonitrile) was added. Additional methanol was added to dip the soil completely. The soil suspension was mixed well using a rotary agitator (Unimax 1010, Heidolph, Germany) for 1 hour and then left at room temperature for 24 hours to allow complete solvent evaporation. After the complete evaporation of the solvent, the fortified soil was again mixed and then serially diluted with untreated soil to get a 5 and 0.5 mg/kg level of fortification. For 5 mg/kg treatment, 1350 g untreated soil was taken in a polythene bag and 150 g of fortified soil (50 mg/kg) was added and thoroughly mixed for homogeneity. For the 0.5 mg/kg treatment, 5 mg/kg treated soil was mixed with untreated soil in the ratio 1:9. The treated soils were maintained at 60% of the maximum water holding capacity (WHC) and stored in a dark at room temperature (20±1C). Moisture contents were maintained at a constant level throughout the experiment by adding distilled water as necessary. Three parallel soil samples (25 g) at both concentration level including unspiked controls were used for analysis of IMI residues and its metabolite (6-CNA) at intervals of 0, 7, 15, 30, 45, 60, 75, 90, 105, 120, 150 and 180 days after application.

## **3.5 Extraction of IMI and 6-CNA from soil samples**

498 Pesticides in the Modern World - Risks and Benefits

solute concentrations, therefore no correction was required. Control samples, containing no IMI, only soil and 0.1 M calcium chloride, were used for each series of experiment. The amount of IMI sorbed to soil after equilibration was calculated from the difference between

*m*

*sor* is the amount of IMI sorbed at equilibrium (mg/kg), *m* is the mass of soil (g), *γ<sup>i</sup>* is

the initial concentration of IMI (mg/L), *γ<sup>e</sup>* is the equilibrium concentration of IMI (mg/L), *V* 

Desorption experiments were conducted on triplicate soil samples immediately after the sorption experiments with the same initial concentrations of IMI. After completing the sorption process, the supernatant (25 mL) were removed and replaced with the same volume of 0.01 M calcium chloride and 100 mg/L mercury chloride solution. After shaking for 24 h, the suspensions were centrifuged under the conditions described previously, and the concentration of IMI was determined in the supernatants using the HPLC. This desorption procedure was repeated five times for each soil sample. The amount of pesticide remaining sorbed by the soil was calculated as the difference between the equilibrium

> *des sor des e ee <sup>V</sup> q =(<sup>γ</sup> - <sup>γ</sup> )*

> > *5 des e*

∑

*sor e*

The persistence of IMI in the tested soils was studied at two concentration levels, 0.5 and 5 mg/kg under laboratory conditions at room temperature (20±1C). For fortification of the soil at 50 mg/kg level, 100 g weighted, air-dried and sieved soil was taken in a baker and 10 mL of standard solution of IMI (1 mg/mL, in acetonitrile) was added. Additional methanol was added to dip the soil completely. The soil suspension was mixed well using a rotary agitator (Unimax 1010, Heidolph, Germany) for 1 hour and then left at room temperature for 24 hours to allow complete solvent evaporation. After the complete evaporation of the solvent, the fortified soil was again mixed and then serially diluted with untreated soil to get a 5 and 0.5 mg/kg level of fortification. For 5 mg/kg treatment, 1350 g untreated soil was taken in a polythene bag and 150 g of fortified soil (50 mg/kg) was added and thoroughly mixed for homogeneity. For the 0.5 mg/kg treatment, 5 mg/kg treated soil was mixed with untreated soil in the ratio 1:9. The treated soils were maintained at 60% of the maximum water holding capacity (WHC) and stored in a dark at room temperature (20±1C). Moisture contents were maintained at a constant level throughout the experiment by adding distilled

*q P = ×100 q*

*des a=1*

*des* is the amount of IMI remaining sorbed by the soil (mg/kg), *γ<sup>e</sup>*

of IMI (mg/L). The percentage of IMI desorbed was calculated as follows:

*m*

(9)

(10)

(11)

*des* is the equilibrium desorption concentration

*sor* is the equilibrium

the initial and equilibrium solution concentration using the mass-balance equation:

*sor e ie <sup>V</sup> q =(<sup>γ</sup> - <sup>γ</sup> )*

is the volume of the solution (L) from which sorption occurs.

sorbed and the desorbed amount by the following equation:

sorption concentration of IMI (mg/L) and *γ<sup>e</sup>*

**3.4 Degradation experiments** 

where *qe*

where *qe*

IMI and 6-CNA were extracted from soil samples according to the method of Baskaran et al. (1997). At each sampling time, a 25 g sample of spiked and homogenized soil was extracted with 40 mL of acetonitrile-water (80:20, v/v) and shaken vigorously for 2 h using a rotary agitator (Unimax 1010, Heidolph, Germany) at 20 (±1)°C. After this time every sample was centrifuged for 20 min at 6000 rpm (BR4i Multifunction, Thermo electron corporation, France) and filtered through a polypropylene hydrophilic filter of 0.45 µm (Whatman, Puradisc 25 TF, USA). The operation of shaking and filtration was repeated three times and supernatants from each extraction were pooled. The solution was evaporated to dryness on a rotary evaporator (Laborota 4002/03 Control, Heidolph, Germany). The residue was dissolved in 1 mL of mobile phase (acetonitrile-water, 20:80, v/v). Three replicates of both level, including unspiked controls, were extracted and analyzed by HPLC.

#### **3.6 Analysis of IMI and 6-CNA by HPLC**

The concentration of IMI and 6-CNA in aqueous solutions was determined using a reversephase HPLC system (Thermo Separation Products, Spectra System, USA) equipped with a UV/VIS detector. All analyzes were performed on a Supelco reverse phase C18 column (150 mm length, 46 mm ID, 5 μm particle size). The mobile phase of acetonitrile and water (20:80 v/v) was used under isocratic conditions at a flow rate of 1.2 mL/min. The analytes were analyzed at 270 nm wavelength. The injection volume and the column temperature were 20 μL and 25 C, respectively. Under these conditions the retention times of IMI and 6-CNA were 4.3 and 1.6 min, respectively.

Calibration curves for both of chemicals were linear from 0.05 to 10 mg/L with regression coefficients of *R2* > 0.999 (six calibration points, in triplicate). The detection limits of IMI and 6-CNA were 0.001 mg/L and 0.003 mg/L, while the lower limits of quantification (LOQ) were 0.005 mg/L and 0.01 mg/L. The mean recoveries for IMI and 6-CNA were 91.4% and 87.8 % with a relative standard deviation lover than 5%.

#### **3.7 Statistical analysis**

Relationship between soil properties and sorption as well as degradation behavior of IMI was tested by a nonparametric correlation test, Kendall-Tau. Except nonparametric tests, multiple linear regression analysis was used, which combines the relationship between different soil parameters and the sorption coefficients, as well as DT50, allowing the assumption of linear models for these parameters (Boivin et al*.*, 2005). Differences in the soil sorption capacity among and within regions were analyzed using Mann-Whitney U test, while the DT50 values were tested by one-way ANOVA test with *post hoc* comparison (Tukey HSD test) to determine the effect of initial concentration and soil on the DT50 of IMI. Data are reported as mean ± standard deviations. The results were considered statistically significant at *p* < 0.05. The data were analyzed using Statistica® software package Version 7.0 and Wolfram Research Mathematica® software package Version 7.0.

Behavior and Fate of Imidacloprid in Croatian Olive Orchard Soils Under Laboratory Conditions 501

**Imidacloprid sorbed (mg/kg)**

**b) Imidacloprid desorption isotherms**

**Imidacloprid in solution (mg/L)** 0.0 0.2 0.4 0.6 0.8 1.0

> Krk I, experimental (mean ± stdv) Istria I, experimental (mean ± stdv) Krk II, experimental (mean ± stdv) Istria II, experimental (mean ± stdv) Krk I, Freundlich model (*R2*

Istria I, Freundlich model (*R2*

Krk II, Freundlich model (*R2*

Istria II, Freundlich model (*R2*

Desorbed amount*<sup>b</sup>* (mg/kg)

17.94 35.96 4.15 23.12

9.36 36.00 1.53 16.39

4.63 37.13 0.47 10.04

1.91 37.88 0.24 12.70

 = 0.988 )

> = 0.982 )

 = 0.989 )

> Desorbed amount (%)

 = 0.997 )

**a) Imidacloprid sorption isotherms**

**Imidacloprid sorbed (mg/kg)**

0

model.

Istria II

Istria II

Istria II

Istria II

concentration.

desorption reaction time.

Soil Initial

concentration, γ (mg/L)

10

5

2.5

1

5

10

15

20

**Imidacloprid in solution (mg/L)** 0 2 4 6 810

Istria I, Freundlich model (*R2*

Krk II, Freundlich model (*R2*

Istria II, Freundlich model (*R2*

Krk I, experimental (mean ± stdv) Istria I, experimental (mean ± stdv) Krk II, experimental (mean ± stdv) Istria II, experimental (mean ± stdv) Krk I, Freundlich model (*R2*

 = 0.996 )

> = 0.998 )

= 0.997

)

Freundlich model. Values are means ± standard deviations. Symbols represent the

Sorbed amount*<sup>a</sup>* (mg/kg)

Fig. 6. a) Sorption and b) desorption isotherms of IMI in the tested soils represented by the

experimental data, while lines represent the theoretical curves described by the Freundlich

Krk I 9.21 18.45 4.84 52.57 Istria I 15.17 30.41 4.20 27.71 Krk II 8.23 16.50 4.84 58.75

Krk I 5.17 19.89 2.35 45.46 Istria I 8.13 31.26 1.84 22.58 Krk II 4.59 17.65 2.52 55.01

Krk I 2.63 21.06 1.25 47.60 Istria I 4.06 32.54 0.97 23.90 Krk II 2.29 18.34 1.22 53.27

Krk I 1.19 23.68 0.49 41.05 Istria I 1.66 33.01 0.25 15.12 Krk II 0.99 19.75 0.44 43.98

*<sup>a</sup>* sorbed amount of IMI after 48 h of sorption reaction time; *b* desorbed amount of IMI after 144 h of

Table 2. The sorbed and desorbed amount of IMI in the tested soils in relation to the initial

All sorption data fit the Freundlich equation (*R2* > 0.966) and Table 3 summarizes the sorption capacity (*KFsor*) and intensity (*1/nsor*) values. The *KFsor* values obtained from the Freundlich

Sorbed amount (%)

 = 0.998 )
