Prunus dulcis shells as precursors for activated carbon were purchased from a local market in Rawalpindi Pakistan. Analytical standard carbendazim, methyl 1- H- benzimidazole-2-carboxylate (MBC), sodium chloride, calcium chloride and potassium bromide were purchased from ACCU Standard, USA, and were used without further purification. Analytical grade solvents, acetone and methanol used were 99.9% pure and were used in solution preparation and apparatus sterilization. A standard stock solution of MBC was prepared in distilled water (DW). Concentrated H₂SO₄ was used for activation of Prunus dulcis shells. Weighting balance (ATx224, Shimadzu company), weighting machine (AUX220, Shimadzu company), orbital incubator shaker (Irmeco Gmbh Germany), centrifuge (Sigma 26-E, Hettich Company), hot plate (MSH-20D, Wisestir instruments), atomic spectrophotometer (220, Varian company), UV-visible spectrophotometer (BMS-1602), octagonal sieve shaker (Endecotts company), pH meter (WTW Ino Labs company) and EC meter (Crison company) were used in the experiments.

Soils from six different regions were used like adhesive media for the MBC molecules to form a thin film (TF) and thin film detachment (TFD) from five provincial territories of Pakistan i.e. Ormara (Balochistan), central Gilgit (Gilgit Baltistan), Pasni (Balochistan), Mianwali and Sialkot (Punjab) and Swabi (Khyber Pakhtunkhwa) (Fig. 1). Sampling for adhesive media was done in May 2016 from regions where no pesticides have been applied, so that the behavior of MBC can be assessed. Samples were taken from the plough layer at depth of 0-6 cm. These soils were shifted to clean polyethylene bags and tightly sealed and transferred to the laboratory for further formulations to be used in sorption and remediation assays. The soils were air dried for 48 hours in a greenhouse at ambient temperature, follow by thorough grinding with pestle and mortar. Prepared samples were sieved with a 2 mm sieve individually and mixed thoroughly until homogeneous. Homogenized soil samples were then used for the determination of physical and chemical characteristics in addition to heavy metals determination (Table II). Samples were stored in sterilized bags for further processing.

Fig. 1

TABLE II

Eq. No	Equations for adsorption-desorption	Description
1	Cs = V / m× (Cb - Ca)	Cs = MBC concentration adsorbed V = solution volumetric extent m = selected soils quantity in g Ca = equilibrium concentration of the supernatant Cb = the equilibrium concentration of the blank
2	Kd = Cs / Ce	Kd = linear adsorption coefficient in µg/ mL Ce = concentration of the MBC at equilibrium in µg/mL Cs = concentration of MBC adsorbed in µg/mL
3	Kf = Cs / Ce	Kf = Freundlich adsorption coefficient in µg/ mL Ce = concentration of the MBC at equilibrium in µg/mL Cs = concentration of MBC adsorbed in µg/mL
4	C = k f · C e 1/n	1/n = Freundlich constant obtained by using the Freundlich equation in linear form
5	K oc = K d /%C · 100	Koc determination is necessary for mobility determination. The values of Koc were checked according to McCall classification and mobility extents were predicted
6	K foc = K f /%C · 100	Similar to Koc, Kfoc determines the MBC affinity for thin film formation on selected soils
7	K OM = K d /%OM · 100	Coefficient for organic matter determination in relation to Freundlich adsorption coefficient
8	LogCs = LogKf + 1/n LogCe	Kf is Freundlich adsorption coefficient, 1/n is Slope of Freundlich adsorption isotherm, n is a linearity factor, also known as adsorption intensity and log Kf is the intercept of the straight line resulting from the plot of log Cs versus log Ce while Cs and Ce are defined previously
9	∆G = -RTLnKOm	R= universal gas constant T= temperature in Kelvin Ln= Log natural
10	H = nads /ndes	nads and ndes ratio for Freundlich adsorption and desorption constants

The specific batch equilibrium method for adsorption desorption of MBC in six selected soils, test guideline 106 given by the Organization for Economic Co-operation and Development (OECD), was followed. Adsorption and desorption experiments were performed at ambient temperature (20-25ºC). An MBC 10 ppm stock solution was prepared by dissolving 10.3 mg of MBC in DW with a few drops of acetone and stirred for 24 h for ensuring the complete dissolution. This was followed by dilutions preparation from MBC stock solution, i.e. 0, 0.25, 0.5, 0.75, 1, 2.5, 5 and 7.5 ppm in 15 mL test tubes, filled with 0.1 M NaCl as background electrolyte for adjustment of ionic strength. Prepared samples aimed at MBC thin film formation on the targeted soils in addition to their duplicates and blanks were shaken for 48 h at 150 rpm on a rotary shaker. Suspensions were centrifuged at 3500 rpm for 10 min. This step developed equilibrium between soil solutions and MBC. Absorbance after filtration using a 0.2 µm membrane filter was spectrophotometrically determined with a UV-Visible spectrophotometer (BMS-1602). The clear aliquots were taken for the analysis.

Desorption experiment for assessing the detachment pattern of MBC thin film was done immediately after the adsorption experiment by decantation of the supernatant from the reaction vials. 9 mL of freshly prepared CaCl₂ was poured into the decanted tubes in order to compensate for possible losses occurring during the reaction. Reactions tubes were shake on a rotary shaker for 48 h at 150 rpm and centrifuged at 3500 rpm for 10 min, then filtered with a 0.2 µm membrane filter and the clear aliquots were taken for the analysis.

UV-Vis data obtained for MBC TF formation and TF detachment was stored in form of absorbance and Lambda maximum on the program associated with UV-Vis (BMS-1602) in the range of 200-600 nm. MBC concentration dependent data were then used for plotting the linear and Freundlich adsorption isotherms for investigating the multilayer adsorption-desorption of MBC on selected soils through different coefficients, e.g. K_d, K_oc, K_f, K_foc etc. Calculations for adsorption and desorption data were performed following the equations of table II.

Selected soils in the investigation expressed distinctive physical and chemical characteristics related to their spatial and climatic variability. As shown in table III, soils from different provincial regions expressed variable traits. The pH ranged between 6.6 to 8.6 for all soils, with the highest pH for soil sample 1 (Ormara, Balochistan). Alkaline pH range can either be attributed to the presence of parent material having higher calcareous composition or it can be a result of various processess, e.g. liming of soils by local farmers. The lowest pH range corresponds to soil sample 3 (Pasni), which is also located in the same provincial territory, but due to the closeness of sea shore, soils are rich in acidifying salt content, thus reducing the pH range. Selected soils also expressed variability in SOM ranging from 1.2 to 3.3 %. As Pakistani soils are not abundant in SOM as a whole, the presence of 3.3 % in the soil sample 6 (Swabi, KPK) is suggestive of higher decomposition rates wither triggered by a large microbial community or by the virtue of climatic conditions. Furthermore, soils were also different from each other for textural characteristics. In terms of the current interest for the adsorption-desorption process, soil sand content and soil clay content are of the prime significance. Soil sample 4 (Mianwali, Punjab) exceeded all soils in possessing higher sandy particles due to geographical attributes, while soil sample 6 (Swabi, KPK) was found to be rich in clay content among all tested soils. Heavy metals such as Pb, Cd, Ni, Cu and Zn were also determined in the selected soils. High levels of heavy metals can be a source of toxicity not only for the crops grown in that region, but also for the soils, and may leach down to soil profiles and underground water storage bodies.

TABLE III

Soils	Location	pH	SOM %	Total nitrogen %	TOC %	Sand %	Silt %	Clay %	CEC (Meq 100/g)	EC (µs/cm)	Pb ppm	Cd ppm	Ni ppm	Cu ppm	Zn ppm
S1	Ormara	8.6	1.2	0.34	0.69	19	74	7	6.7	272	2.88	0.05	0.68	0.17	0.07
S2	Gilgit	7.0	1.6	0.39	0.93	54	38	8	7.0	201	1.15	0.09	0.73	0.13	0.30
S3	Pasni	6.6	1.8	0.31	1.04	50	33	17	7.8	84	3.31	1.20	0.73	N.D	0.08
S4	Mianwali	8.1	2.3	0.36	1.33	59	34	7	7.5	120	3.47	0.06	0.68	N.D	0.09
S5	Sialkot	8.0	2.8	0.08	1.62	57	31	12	7.1	189	1.43	0.11	0.69	N.D	0.16
S6	Swabi	7.3	3.3	0.09	1.91	34	48	18	7.4	232	0.72	0.15	0.43	0.04	0.08

*SOM = soil organic matter, TOC = total organic carbon, CEC = cation exchange capacity, EC = electrical conductivity

MBC thin film formation and detachment were analyzed via isotherm based MBC mass transfer. A division between different compartments in a typical sorption behavior was observed. MBC adsorption to the selected soils is nonlinear, confirmed via irreversible adsorption. Two models adopted for this analysis were the linear and Freundlich models. The linear adsorption model was obtained by plotting MBC equilibrium concentration (µg/mL) on the X-axis and MBC adsorbed (µg/mL) on the Y-axis. Various other parameters were derived (Table IV; Table V).

TABLE IV

Soils	Kd(ads) (µg/mL)	R2	Koc	∆G (Kj/mol)	Kf (µg/mL)	R2	Kfoc	na
S1	6.3	0.99	298.0	-9.9	4.8	0.91	513	2.5
S2	7.1	0.98	303.0	-10	5.3	0.99	494	2.3
S3	9.0	0.92	570.0	-19	6.8	0.99	174	1.7
S4	7.3	0.98	343.0	-14	5.8	0.89	492	2.0
S5	9.7	0.68	683.0	-15	6.6	0.88	686	2.7
S6	11	0.73	1007	-15	8.8	0.92	821	1.7

*K_d(ads) = linear adsorption coefficient, K_oc = linear adsorption coefficient normalized for organic carbon, ∆G = Gibbs free energy change, K_f = Freundlich adsorption coefficient, K_foc = Freundlich adsorption coefficient normalized for organic carbon, n_a = Freundlich parameter for adsorption.

TABLE V

Soils	Kd(des) (µg/mL)	R2	Kf (µg/mL)	R2	nd	H
S1	7.4	0.82	6.0	0.99	1.7	0.6
S2	6.7	0.98	5.7	0.91	1.6	0.7
S3	5.2	0.99	3.8	0.92	0.9	0.5
S4	5.6	0.89	4.8	0.89	1.2	0.6
S5	6.3	0.99	6.8	0.79	0.8	0.3
S6	6.5	0.99	6.6	0.83	0.9	0.5

*K_d(des) = linear desorption coefficient, K_f = Freundlich desorption coefficient, n_d = Freundlich parameter for desorption, H = hysteresis

MBC attachment to the selected soils was determined via K_d values, which provided an insight to the level and adsorption potential of each soil. K_d values ranged between 6.3 and 11 µg/mL, with soil sample 1 expressing the lowest MBC adsorption and soil sample 6 adsorbing MBC in greater quantities (Fig. 2). Greater K_d values signify stronger surface attachment of MBC molecules to the soils and makes a thin film over the surface, while the lower values correspond to insignificant attachment on the soil particles. MBC also expressed variable adsorption affinities towards the six types of soil and the results are consistent with other fungicides, e.g. strobilurin fungicides (Azarkan et al. 2016). In the adsorption extent characterization used in Test Guidelines for Environmental Safety Assessment for Chemical Pesticides (USEPA 2008) the K_d values express the adsorption extent, i.e. there is greater difficulty in pesticide adsorption if K_d < 5, comparatively less difficulty if K_d is 5~20 and finally pesticide adsorption will be of medium scale if K_d is between 20~50. The application of this criteria to the obtained results shows that the adsorption of MBC by the selected soils was not difficult, since the K_d range was 6.3 - 11 µg/mL, namely a type II adsorption level for all soils. A similar trend can also be observed for K_f (Fig. 3).

Fig. 2

Fig. 3

MBC adsorption variability checked via K_d and K_f can be comprehended in terms of the MBC variable response towards different soils for the formation of a thin film. The most significant factor in this regard is the SOM. SOM in the current investigation varied from 1.2 to 3.3 %. SOM provides different functional groups for the attachment of pesticide molecules such as carboxyl, phenyl, hydroxyl, carbonyl and methyl groups, among others. Thus, SOM act as an attachment media for MBC molecules via production of hydrogen bonds. Soil sample 6 is consistent with this fact. Results for other studies on azoxystrobin also confirm that its dissipation was faster in soils with lower organic matter content than the soils which were deliberately amended with organic content (Herrero-Hernández et al. 2015) MBC residence in soils is thus dependent upon the MBC properties, soils’ physical, chemical and biological characteristics, soil organic amendments etc. MBC is water soluble, thus, there are chances of its leaching to groundwater resources as it did not express strong adsorption relationship with the soil particles or SOM, and prevented surface entrainment (Pose-Juan et al. 2015). In addition to the SOM, clay content of soils also influence adsorption extent. Pakistan has variable soil textures due to sharp difference between the weather pattern, climatic variability and parent rock material. In this finding, soil sample 6 having 18 % clay content expressed greater adsorption. This can be attributed to the composition of soil sample 6, being the smectite family of clay minerals and consequently expressing greater retention of MBC. Furthermore, the layered sheet like structure of phyllosilicates has negative charges compensated by positive charges in the spaces of the interlayers. Such complicated layered morphologies not only provide a higher surface area for MBC molecules but also develop hydrogen bonding with it. In addition to SOM and clay content, soil pH also has great impact on the level of adsorption. Current results are suggestive of better adsorption in soils having lower pH and decreased adsorption in soils having higher pH. Current results are consistent with other Benzimidazole fungicide,s e.g. thiabendazole (de Oliveira et al. 2017)

For other pesticides, e.g. atrazine, isoproturon and paraquat, results are also suggestive of little impact of dissolved organic matter on adsorption. Such behavior can be best comprehended by developing a correlation analysis between different parameters of soil and MBC adsorption. Such an approach will also enhance the understanding of soil clean up strategies (Liu et al. 2018). MBC molecules detachment from selected soils is dependent upon soils’ physical and chemical characteristics (Pavlovic et al. 2018). Soil sample 1 expressed highest desorption potential for having the highest pH, 8.6, and lowest organic matter, 1.2 %. Comparatively easier MBC TF detachment from soil sample 1 is suggestive of the ease leaching of the MBC molecules in its intact form or as converted daughter products towards the lower soil profiles and underground water reserves. Furthermore, these residues might be taken up by crops grown in the following season and might also be translocated to regions through volatilization, leaching, etc. In this research, it can be seen that n_a value of soil 6 is lesser as compared to the remaining five soils. It indicates the intensity of adsorption. Similarly, the n_d values are used to check the intensity of desorption occurring in the soil samples. In the present study desorption hysteresis coefficient, H, in all six soils is found to be from 0.3 to 0.7, which demonstrates that hysteresis is present. Although the occurrence of hysteresis indicates that the adsorption isotherms are slightly different from the desorption isotherms, the value of H close to 1 means that desorption process took place almost as quickly as adsorption did. The value of hysteresis is closer to 1 in soil 2 indicating that the soil have undergone the process of desorption almost as quickly as adsorption. This effect can be justified by the fact that the sand content in soil 2 is quite high, resulting in the process of desorption almost as quickly as adsorption.

MBC percolation to underground water bodies is an important factor for polluting the resources, which cannot be cleaned in a simple way. MBC percolation in soils can be determined from different techniques, e.g. thin layer chromatography (TLC) or regression analysis. As predicted by K_d values showing difficult adsorption, the values for K_oc are also suggestive of medium mobility of MBC to soil profiles and water bodies present underground in terms of McCall classification (McCall et al. 1980). This is suggestive of the fact that failure in timely remediation of MBC in the selected soils may result in contamination that is difficult to control later on. Particularly the absence of SOM and lower clay contents enhance this percolation. Current results are in conformity with the previously studied fungicides, e.g. tebuconazole (Vanclooster et al. 2000, Singh et al. 2016). Other investigations on hydrophobic fungicides are indicative of lower mobility patterns as in the case of pyraoxystrobin (Liu et al. 2018). In this regard, various researchers have also employed ground ubiquity score (GUS) and suggested that GUS < 1.8 ensures minimum leaching (Wu et al. 2016). The current investigation of MBC adsorption in selected soils provides important information about its behavior in soils.

MBC sorption in soil is determined via adsorption free energy (ΔG). Current results have given negative values for ΔG for MBC adsorption into the soils, consistent with other fungicides, e.g. 14C-pyraoxystrobin (Liu et al. 2018) and strobilurin fungicides (Wu et al. 2016). A negative value of ΔG for MBC adsorption shows that MBC molecules were not tightly bound to the interlayer of soils selected for adsorption, and that they were held together by temporary weaker bonds, e.g. hydrogen bonds, Van der Waal’s forces etc. However, in case of strong adsorption, referred as chemisorption, MBC would have yield positive ΔG values. Such behavior of MBC is also suggestive of adsorptive interactions taking place only in the surface soils and that it has not invaded the deeper profiles (Liu et al. 2018). Negative values of ΔG can be converted to positive values through organic amendment of selected soils, which will firmly held the MBC molecules (Sharma et al. 2015).

Statistical investigations were done regarding the relationship between K_d and the physical and chemical properties of soil including OM, TOC and pH. Linear regression analysis was performed to analyze the effect on K_d of the physical and chemical properties (Fig. 4, Table VI). The analysis provided the knowledge that pH is negatively correlated to K_d, while OM and TOC were seen to be positively correlated. According to figure 4, the lowering pH enhances the adsorption of MBC in soils. In a nutshell, soil pH values are accountable for the dissociation or protonation processes of both, the MBC and the adsorbent surfaces of soils. Whereas, an increase in adsorption was observed with the increasing percentage of OM and TOC. The results also indicated that the soil 6, having the highest K_d(ads) value (11 µg/mL), contains the most percentage of soil OM (3.3 %), proving the fact that Soil OM is directly proportional to the rate of adsorption occurring in that particular soil. Also, the CEC value of 7.4 indicates that comparatively higher CEC values usually enhance adsorption, either by ion exchange or surface precipitation. One-way analysis of variance (ANOVA) was performed on the soil samples and their K_d values (Table VII). Different parameters of ANOVA included mean square (MS), P value (probability), degrees of freedom (df), sum of squares (SS), F statistics (F) and F critical values (F crit). The P value (0.9) was found to be greater than the alpha value (α) (0.05). The F crit value was found to be 3.0 and the F statistics value was lower than it (0.2). Value of F statistics less than the F crit value proves that the experimental values are accurate and in an acceptable range. Furthermore, residual plots and normal probability plots were obtained in order to observe the goodness of fit of the experimental results (Fig. 5). Normal probability plots depict that the data is distributed normally while the residual plots display a constant variance of the data.

Fig. 4

TABLE VI

Sorption coefficient	Property (x)	Correlation coefficient (r)	Probability level (p)	Intercept (a)	Slope (b)
Kd	pH	-0.3	0.05	8.8	-0.1
	SOM	0.9	0.06	1.3	0.6
	TOC	0.8	0.02	-0.6	0.2

*K_d = adsorption coefficient, SOM = organic matter, TOC = total organic carbon

TABLE VII

Source of variation	SS	df	MS	F	P-value	F crit
Between soil samples	10.0	5.0	2.0	0.2	0.05	3.0
Within soil samples	149	12	12

MS = mean square, P value = probability, df = degrees of freedom, SS = sum of squares, F = F statistics, F crit = F critical values.

Fig. 5

PD-AC was used for the remediation of the MBC contaminated soils in three different concentrations (2.5, 5.0, and 7. 5 ppm) of MBC. Solutions having different initial concentrations (2.5, 5.0, and 7. 5 ppm) of pesticide were measured for the absorbance using a UV-Vis spectrophotometer. The standard plots were obtained with “absorbance versus initial concentrations” on MS Excel 2010 (Fig. 6). After approximately 3 h, all of the MBC was removed by PD-AC from the 2.5 ppm solution (100 %; Table VIII). While 36 % removal was observed from 5 ppm solution and 23 % removal from 7.5 ppm solution. It can be deduced from the obtained removal percentages that the amount of MBC removed from a solution by PD-AC depends highly upon the concentration of MBC used. This shows an inverse relation between MBC removal and concentration. The lower concentration of solution means there are lesser molecules of MBC in solution, hence the percentage of removal is high as it takes less time for those molecules to get adsorbed on the surface of the activated carbon. Consequentially, a declining trend in pesticide removal with increasing concentration was observed. The use of PD-AC rendered this process to be environmentally friendly by using organic materials instead of hazardous chemicals. This process proved to be quite cost effective and economical also aiding in waste management by reusing wasted almond shells and requiring very low financial input while in turn providing a significant output.

Fig. 6

TABLE VIII

Solution concentration (ppm)	UV absorbance before adding activated carbon	UV absorbance after addition of activated carbon	% removal of pesticide from solution
2.5	0.090	0.090	100
5.0	0.324	0.115	36
7.5	0.498	0.116	23

*UV = ultraviolet