Movement and Fate of 2,4-D in Urban Soils: A Potential Environmental Health Concern.

The fate and movement of 2,4-dichlorophenoxyacetic acid (2,4-D), in terms of sorption-desorption and leaching potential, were evaluated in urban soils following the batch experimental method. The sorption kinetics of 2,4-D in soils followed both "fast" and "slow" sorption processes that could be well described by a pseudo-second-order kinetics model, suggesting that 2,4-D was partitioned into soil organic matter and clay surfaces, and eventually diffused into soil micropores. The sorption isotherms were linear, following both Langmuir and Freundlich models. Partially decomposed or undecomposed organic matter present in urban soils decreased sorption and increased desorption of 2,4-D. Also, sorption of 2,4-D increased with an increase in the contents of clay and Al and Fe oxides, whereas sand and alkaline pH increased the desorption process. The lower calculated K d values suggest that 2,4-D is highly mobile in urban soils than in agricultural soils. The calculated values of groundwater ubiquity score, leachability index, and hysteresis index indicated that the herbicide is highly prone to leach out from surface soil to groundwater which might affect the quality of potable water. The present study clearly suggests that 2,4-D must be judiciously applied in the urban areas in order to minimize the potential health and environmental risks.

Introduction

In the urban environmental settings such as lawn/turf, home gardens, and parks, application of herbicides is a more common feature almost all over the world. In fact, the use of pesticides by homeowners in the urban environment, on per acre basis, is 10 times more than the use on crops by farmers.[1] When pesticides are applied on lawns, trees, shrubs and flowering plants, both short- and long-term health problems result in many people, animals, insects, wildlife, and fish.[2] The toxic effects of 2,4-D in non-target organisms depends on several factors including the frequency of exposure, dose, and the host factors which influence susceptibility and sensitivity.[3] Thus, in the urban environment, the consequences due to injudicious and indiscriminate application of herbicides are severe, as countless children nationwide play on lawns in schools and parks and at homes.[1,2]

2,4-Dichlorophenoxyacetic acid (2,4-D) is among the most frequently used synthetic active ingredients of plant protection products worldwide.[4] It is the first commercially developed selective herbicide, applied in the fields grown with cereals, lawns, gardens, and vineyards[5] for killing dicots without affecting monocots, as it mimics the natural auxin at the molecular level.[6] 2,4-D was registered for its use in Australia more than 50 years ago, and currently, there are about 220 products containing 2,4-D for their use in Australia (http://apvma.gov.au/node/15581). As per the report of US EPA,[7] the annual use of 2,4-D in urban agricultural activities and non-agriculture settings in the USA alone was 30 and 16 million pounds, respectively. Although there are several formulations of 2,4-D including acids, esters, and salts which vary in their chemical properties, after application into the soil, these compounds get converted (hydrolyzed or degraded) rapidly to the acid form.[8,9] This moderately persistent herbicide is most commonly detected in streams and surface and groundwaters worldwide.[10] Likewise, 2,4-D used in households is frequently detected in urban waterways in Australia, Canada, Europe, and USA.[11,12] 2,4-D has been found in urban surface waterways at a concentration of 3.5 ng L–1 in Australia.[13] Also, the report of a community study indicated that 2,4-D was detected in urine of pet dogs exposed to household lawns treated with this herbicide in the USA.[14] 2,4-D is one of the well-known endocrine-disrupting chemicals having a serious impact on the endocrine and immune system.[15] Although the International Agency for Research on Cancer classified 2,4-D as a possible human carcinogen and mutagen,[16] it is still being extensively used worldwide both in agricultural and urban environments. Therefore, the study on the movement and fate of 2,4-D acid form in urban soils is critical.

Potential contamination of pesticides occurs by leaching into groundwater through surface runoff, plant uptake, volatilization, and drift through the air.[1] In fact, the fate and movement of the pesticides in diverse sections of the environment are strongly influenced by sorption and desorption, as these are the fundamental processes that influence their mobility and bioavailability.[17] A large body of information on sorption–desorption of 2,4-D in agricultural soils is available in the literature;[10,18−21] however, virtually no reports are available on urban soils, although such data are very much essential to understand the potential health risks of pesticides in humans. The changes in urban soils that occur due to infrastructure development result in pervious and impervious surfaces (footpaths, streets, concrete surfaces) with micropore collapse, dense particle packing, ped breakdown, high bulk density, and so forth.[1] are likely to enhance the movement of a contaminant on the soil surface and into groundwater. Moreover, approximately 90% of the urban surface areas comprise impermeable structures, and pesticides are the direct sources of contamination in urban waterways.[1] Because urban soil properties differ significantly from any other natural soils and are more strongly influenced by anthropogenic activities,[1,22] an understanding of the fate and behavior of 2,4-D in the urban environment is much warranted. Therefore, the present study investigates for the first time the sorption–desorption behavior and health risk potential of 2,4-D in nine urban soils of varying physicochemical characteristics, collected in and around Newcastle, Australia.

Results and Discussion

Sorption of 2,4-D in Selected Urban Soils

The urban soils used in the present investigation belonged to six distinct textural classes such as loam, sandy loam, silt loam, sand, loamy sand, and clay (Table ). In all soils, quartz was the major mineral component followed by birnessite, albite, zeolite, sodalite, and so forth. The soil pH, electrical conductivity (EC), and cation exchange capacity (CEC) were in the range of 5.5–8.0, 42.6–661 (S m–1), and 4.22–10.01 meq 100 g–1, respectively. The total organic carbon (TOC), aluminum (Al), and iron (Fe) in soils were in the range of 0.19–7.66, 0.01–1.27, and 0.03–3.15%, respectively (Table S1). The ranges of nitrogen (N), sulfur (S), calcium (Ca), potassium (K), magnesium (Mg), and sodium (Na) of all investigated soils were, 0.03–0.46, 0.01–0.14, 0.03–1.05, 0.01–0.17, 0.12–0.30, and 0.01–0.17%, respectively (data not shown).

Table 1

Physicochemical Properties of Soils Collected from the Urban Environment In and Around Newcastle, Australiaa

soil ID	soil collection area	clay (%)	silt (%)	sand (%)	textural class	major mineral compound	TOC (%)	Al (%)	Fe (%)	pH (in Milli-Q water)	EC (μs cm–1)	CEC (meq100 g–1)
ATC	UoN-1	7.5	41.2	51.3	loam	Quartz, Orthoclase, Albite, Hyalophane	7.66	0.73	1.10	5.8	117.6	5.35
TAR	Taree	11.2	55	33.8	silt loam	Quartz, Sinnerite, Ice Ic, Sylvine, Bernalite, Albite	2.02	0.92	1.19	7.5	161.8	4.22
MAR	Maryland	7.5	16.2	76.3	loamy sand	Quartz, Albite, Zeolite, Sodalite	0.21	0.73	3.15	8.0	285.5	9.23
SAL	Salamander Bay	1.2	1.2	97.6	sand	Quartz, Dolomite, Zeolite LC-3, Palladium	0.25	0.65	1.42	6.1	42.6	6.37
WAB	Warabrook	18.7	25	56.3	sandy loam	Quartz, Birnessite, Albite, Sinnerite	1.44	0.96	1.40	6.6	107.0	7.91
WAT	Waratah	30.0	41.2	28.8	loam	Quartz, Birnessite, Zeolite Rho, Albite	0.19	0.81	1.08	5.8	661.0	10.01
NLT	North Lambton	42.5	21.2	36.3	clay	Quartz, Birnessite, Anorthite (sodian), Kaolinite 1A	0.82	0.88	2.11	5.5	116.2	7.62
FOR	UoN-2	7.4	23.8	68.8	sandy loam	Quartz, Marshite, Albite, Zeolite	3.52	1.27	1.95	6.2	123.3	6.63
FLE	Fletcher	12.4	23.8	63.8	sandy loam	Quartz, Oligoclase, Albite, Sodalite	1.29	0.01	0.03	6.6	190.5	9.70

UoN = University of Newcastle.

Sorption kinetics is one of the most significant characteristics leading solute uptake rate and represents sorption efficiency.[23] Batch sorption kinetic studies were performed at a concentration of 10 mg L–1 2,4-D using nine Australian urban soils. Sorption kinetics for 2,4-D exhibited two distinct behaviors (Figure a,b). The initial step was a rapid drop in herbicide concentration indicating a faster sorption mechanism. In the second step, the decline in herbicide concentration continued at a slower rate and finally reached an equilibrium concentration after 3 h. This observation of the two-step process indicates that many vacant surface sites were available for sorption at the beginning. After the vacant surfaces were occupied by 2,4-D molecules with the increase in time, other molecules further migrated through intraparticle diffusion, which is comparatively slower than the surface layer diffusion. Hence, the sorption rate decreased until reaching the equilibrium concentration. Similar findings were reported in soils for endosulfan[24] and atrazine.[25] The extent of 2,4-D sorption at equilibrium time was approximately 58.31, 36.78, 26.00, 20.70, 32.51, 25.40, 31.06, 44.62, and 37.75% for soils designated as ATC, TAR, MAR, SAL, WAB, WAT, NLT, FOR, and FLE, respectively. The correlation coefficient values (R2) of the pseudo-first-order kinetics model were in the range of 0.8544–0.2381 and were lower than those obtained in pseudo-second-order kinetics model which ranged from 0.9998 to 0.9864 (Table S3). Thus, the pseudo-second-order kinetic equation is suitable to model the sorption curves of 2.4-D which indicated that the rate-limiting step was chemical sorption, including electronic forces through sharing or exchange of electrons.[26] Moreover, it suggests that the process is governed by the availability of sorption sites on the soil surfaces instead of 2,4-D concentration in solution.[27]

Figure 1

Equilibrium studies of 2,4-D. (a,b) sorption; (c,d) desorption in nine different urban soils. See Table for details of soil IDs used.

The soils containing higher amounts of organic carbon (OC) exhibited maximum sorption of 2,4-D, indicating the direct influence of OC in sorption of the herbicide. Thus, the soil ATC that contained 7.66% OC (Table ) sorbed 58.31% of 2,4-D added in solution, whereas only 26.00% sorption was evident in soil MAR which had the lowest amount of OC (0.21%). However, the extent of sorption in MAR was more than in soil SAL that contained relatively higher quantity of OC. This may be due to the presence of a greater amount of clay material in MAR (7.5%) than in SAL (1.2%). On the other hand, soil NLT that contained the highest per cent of clay (42.45) and less amount of OC (0.82%) sorbed relatively higher amount (31.06%) of 2,4-D than those soils with a similar or slightly higher quantity of OC. This observation also suggests that OC together with clay plays a vital role for sorption of 2,4-D[19,20] in the urban soil.

It has been reported that not only organic matter and clay content but also soil pH and aluminum-iron oxides play a major role in sorption of 2,4-D in agricultural soils.[18,19,28] As sorption of pesticides is mostly affected by not only the content but also the properties of soil organic matter,[29] well-decomposed black carbon/kerogen facilitate enhanced sorption and very slower desorption.[19] Although the soils selected in the present study contained organic matter at varying levels (0.19–7.66%), partially decomposed or undecomposed organic matter was observed floating in the centrifugation tubes, and presence of this form of OC could be the plausible reason for the lesser sorption of 2,4-D. Furthermore, soils with similar mineral composition and OC exhibited the same pattern of sorption of 2,4-D. For instance, 2,4-D sorption in soils MAR, WAT, SAL, and NLT with a similar mineral composition (quartz) and <1.0% of OC content was nearly the same. Thus, it is clear from the sorption data that soil minerals have a positive role in sorption of 2,4-D, particularly when OC content is below 1%. Similarly, other researchers reported that soil minerals play a significant role when OC content is <1%.[29,30]

Another batch experiment was conducted to explore the sorption isotherms of 2,4-D in nine urban soils applied in the range of 0.5–20 mg L–1. The Langmuir and Freundlich sorption isotherms were applied for quantitative analysis of 2,4-D transport from liquid to a solid phase to understand the nature of interactions between 2,4-D and the soil matrix.[19] The Langmuir isotherm is used mainly for monolayer sorption on perfectly smooth and homogeneous surfaces, whereas the Freundlich isotherm is used to model for multilayer sorption in heterogeneous surfaces.[31] Based on the Langmuir and Freundlich equations, the sorption parameters, qmax, KL, KF, 1/n, and the coefficient of determination (R2) of 2,4-D in test soils were obtained, and the data on Langmuir and Freundlich sorption isotherms of 2,4-D in urban soils are presented in Figure and Table S2. The results indicated that the 2,4-D sorption capacities of the nine tested urban soils varied significantly due to their different physicochemical properties. The sorption of 2,4-D increased with an increase in its concentration. However, at the low concentration of 2,4-D, sorption capacity followed a linear increasing trend. 2,4-D sorption parameters for SAL, WAB, and FLE soils were the best fit in the Langmuir isotherm model with coefficients of determination (R2) 0.9920, 0.9622, and 0.9865, respectively (Figure a,b), which indicated homogenous distribution of active sites on the soil surface. The value of qmax in the Langmuir model denotes the maximum quantity of 2,4-D that could be sorbed by the soils. The observed qmax values (mg kg–1) for 2,4-D in the nine soils (Table S2) followed a descending order: ATC (92.59) > FOR (56.17) > NLT (53.47) > TAR (47.16) > MAR (34.96) > FLE (29.23) > WAB (26.73) > WAT (25.77) > SAL (13.49), and these results were consistent with the Freundlich model. The Langmuir constant (KL) is related to the binding energy of sorption, indicating the affinity of the soil surface for 2,4-D. The KL values for 2,4-D sorption on these soils ranged from 0.07 to 0.43 L mg–1.

Figure 2

Sorption isotherms of 2,4-D in nine different urban soils. (a,b) Langmuir isotherms; (c,d) Freundlich isotherms.

2,4-D sorption data were also well fitted in the Freundlich isotherm model, particularly for soils ATC, TAR, MAR, WAT, NLT, and FOR with R2 values of 0.9882, 0.9978, 0.9819, 0.9823, 0.9865, and 0.9934, respectively (Figure c,d). These data assume that 2,4-D sorption occurs on a heterogeneous surface with the possibility of sorption being multilayered.[19] The present observations on 2,4-D agree with those reported for atrazine in the soils with organic matter and clay content.[32] The calculated values of the Freundlich coefficient (1/n) were <1 (ranging from 0.5398 to 0.8267) (Table S2), indicating that the sorption is by heterogeneous mode wherein the herbicide molecules preferred high-energy sites followed by sorption at lower-energy sites.[33] Moreover, 1/n value < 1 is indicative of favorable chemisorption mechanism on heterogeneous surface.[34] The sorption capacity of the nine urban soils for 2,4-D, as represented by Freundlich sorption coefficient (KF), followed the order: ATC (8.02) > FOR (7.82) > NLT (6.98) > TAR (4.90) > WAB (3.79) > WAT (3.67) > FLE (3.63) > MAR (3.53) > SAL (2.89) (Table S2). The low values of the Freundlich sorption coefficient indicate that 2,4-D is possibly highly mobile in soil.[35]

Overall, the properties and molecular structures of herbicides would affect the degree of sorption in soils. The faster rate of sorption of an herbicide could be attributed to its partition into soil organic matter in anionic and non-ionic form[10] or sorption onto mineral surfaces.[30] In contrast, the slow sorption may be due to the gradual diffusion of organic compounds into soil micropores[30] or into highly cross-linked regions of soil organic matter.[36] In the present study, 2,4-D sorption process in urban soils was dominated by surface sorption at lower equilibrium concentration. With the increase in equilibrium concentration, the surface sorption reached a saturation point and the sorption process was dominated by partition. Also, with an increase in the organic matter and clay content in soils, there was an enhanced sorption of the herbicide. In all soils, quartz was the major mineral component followed by birnessite, albite, zeolite, sodalite, and so forth, and all mineral components had an almost similar positive influence on sorption in all urban soils used. However, an acidic condition was more conducive for sorption of 2,4-D, whereas alkaline soil pH resulted in decreased sorption. Similarly, it has been reported earlier that the extent of 2,4-D sorption was inversely proportional to soil pH.[37] Furthermore, the sorption of 2,4-D onto the urban soil increased with increasing concentrations of metal oxides such as Al and Fe. All these results indicate that organic matter, clay content, acidic pH, Al, and Fe favor 2,4-D sorption in urban soils.

Desorption of 2,4-D in Urban Soils

Desorption plays a vital role in understanding the release rate and the potential mobility of pesticides in soil.[38] Desorption kinetic study for 2,4-D was performed using a concentration of 10 mg L–1 after completing sorption kinetics, and isotherms were determined as described for sorption studies. Desorption of 2,4-D was also initially rapid up to the first 90 min and then continued with a slow rate and reached equilibrium after 180 min (Figure c,d). Desorption coefficient and constant values were obtained from pseudo-first- and pseudo-second-order kinetic models (Table S3). The correlation coefficient values (R2) of the pseudo-first-order kinetics model ranged from 0.0548 to 0.9393, and were lower than those of the pseudo-second-order kinetic equation that ranged from 0.9983 to 0.9999 (Table S3). Therefore, the pseudo-second-order kinetic equation is the best fit in the model for the desorption process.

Desorption isotherms denote the quantity of a contaminant which is still sorbed as a function of the equilibrium concentration after one desorption cycle.[39] Desorption isotherm parameters also fit very well with both Langmuir (Figure a,b) and Freundlich models (Figure c,d) for 2,4-D in the tested urban soils. Particularly, 2,4-D desorption parameters for soils MAR, SAL, and WAT were best fitted in the Langmuir isotherm model, and the coefficients of determination (R2) values were 0.9287, 0.8469, and 0.9658, respectively. The isotherms for soils ATC, TAR, WAB, NLT, FOR, and FLE were best fitted by the Freundlich isotherm model as the R2 values were 0.9844, 0.9619, 0.9891, 0.9972, 0.8933, and 0.9884, respectively (Table S4). On the other hand, the values of KF in all soils tested ranged from 1.67 to 2.76 (Table S4) for the desorption of 2,4-D and were consistently lower than those of sorption KF values that ranged from 2.89 to 8.02 (Table S2). These data indicate the higher desorption capacity as well as higher mobility of 2,4-D in urban soils as reported for this herbicide earlier in Andosol and kaolinite.[40]

Figure 3

Desorption isotherms of 2,4-D in nine different urban soils. (a,b) Langmuir isotherms; (c,d) Freundlich isotherms.

Potential Environmental Health Risks of 2,4-D Sorption–Desorption in Urban Soils

The mobility of herbicide in the environment strongly depends on its sorption–desorption and leaching behavior in soil. The contamination of water bodies with pesticides has been a major concern because it affects the drinking water quality, non-target organisms, and food safety. The leaching potential, represented by Kd and Koc values, is the most crucial factor for the environmental risk assessment of any herbicide.[41] For several agricultural soils that exhibited lesser mobility of 2,4-D, the calculated values of soil/solution distribution coefficients (Kd) ranged approximately from 0.66 to 7.89 L kg–1.[19,21,28] As the calculated Kd values for 2,4-D in the selected urban soils were low ranging from 0.65 to 4.68 L kg–1, this result clearly supports the fact that soils with lower Kd values favor lesser sorption and greater mobility of 2,4-D. The Koc value is frequently used in sorption processes because of the significant role played by the variability of % OC content in soils.[41] In the present study, Koc values (L kg–1) ranged from 771.59 (in soil MAR) to 61.13 (in soil ATC). The half-life of 2,4-D in the urban soils tested was taken as 59.3 days, although this value may vary based on soil depth, microbial activities, and other environmental conditions.[42] Using this half-life of 2,4-D, the values of Koc, groundwater ubiquity score (GUS), leachability index (LIX), and hysteresis index (HI) were calculated (Table ). The GUS values obtained from the equation ranged from 3.92 (in soil ATC) to 1.97 (in soil MAR). Herbicides with GUS values <1.8 are considered as “non-leachers”, whereas those with a value > 2.8 are the “leachers” and others with values between 1.8 and 2.8 are considered “transitional”.[41,43,44] Therefore, 2,4-D has great potential for leaching in eight urban soils studied because their GUS values exceed 2.8, whereas MAR soil is considered as transitional due to its GUS values being between 1.8 and 2.8. The calculated LIX values ranged from 0.01 (in soil MAR and WAT) to 0.48 (in soil ATC). Generally, the LIX index varies between zero and one that represent the minimum and maximum leaching potential, respectively.[41] These results indicate that the soils used in the present study have low to moderate leaching potential as their LIX values are below 0.5. Thus, in the tested soils, 2,4-D may have the potential to contaminate both surface and groundwaters.

Table 2

Environmental Health Risk Assessment of 2,4-D Sorption–Desorption in Urban Soilsa

soil ID	Kd (L kg–1)	Koc (L kg–1)	GUS	LIX	HI
ATC	4.68	61.13	3.92	0.48	0.93
TAR	2.28	112.95	3.45	0.26	0.84
MAR	1.62	771.59	1.97	0.01	0.66
SAL	0.65	263.27	2.80	0.04	0.35
WAB	1.25	87.34	3.65	0.36	1.00
WAT	1.26	665.31	2.08	0.01	0.98
NLT	2.09	256.03	2.82	0.05	0.95
FOR	3.75	106.80	3.49	0.28	0.96
FLE	1.74	135.15	3.31	0.20	0.90

GUS = groundwater ubiquity score; LIX = leachability index; HI = hysteresis index.

HI is a phenomenon that indicates a delay or hindrance in desorption relative to sorption.[45] A calculated HI value close to 1 indicates the absence of hysteresis, whereas a value <1 represents the hysteresis process due to irreversible binding.[45] However, researchers considered any HI value between 0.7 and 1.0 for the absence of hysteresis.[41] In the present study, the HI value ranged from 0.35 to 1.00, indicating that 2,4-D desorption is favored in seven soils tested. On the other hand, in soils MAR and SAL, the HI values are 0.35 and 0.66, respectively, which is indicative of the hysteresis process. Some clay minerals are responsible to make it happen.[41] The irreversible binding takes place in clay minerals, and this is attributed to the OC associated with the soil aggregates and entrapment of sorbed molecules inside meso- and microspores within OC and mineral structures.[45] Seven of the urban soils used have no hysteresis effect wherein 2,4-D desorption is favored. The OC content of four urban soils (ATC, TAR, FOR, and FLE) is comparatively higher than that of the other soils, but the leaching potential, based on GUS, LIX, and HI, was higher which may be due to the presence of partially decomposed or undecomposed OC in urban soils. Thus, the values of GUS, LIX, and HI clearly indicate that 2,4-D is highly mobile in all urban soils used in the present study and are liable to leach out from soil surface to waterways. In addition, 2,4-D from impervious surfaces (approximately 90% present in the urban environment) can easily enter into urban waterways and serve as direct source of contamination. Therefore, contamination of water bodies with 2,4-D applied in the urban environment may pose a potential threat to the aquatic organisms and other living beings as well. In all, this is the first detailed study that evaluated the impact of 2,4-D in urban soils.

Conclusions

The sorption kinetics of the herbicide revealed initial faster sorption followed by slower sorption. The pseudo-second-order kinetic model suggested that 2,4-D partitioned into soil organic matter and clay surfaces initially followed by gradual diffusion of the herbicide into soil micropores. The sorption data for soils SAL, WAB, and FLE were best fitted by the Langmuir model, whereas those for soils ATC, TAR, MAR, WAT, NLT, and FOR by the Freundlich model. Sorption of 2,4-D in urban soils increased with an increase in the contents of organic matter, clay, and Al and Fe oxides, whereas contents of sand and alkaline pH decreased the process. The quality of organic matter in urban soils differ significantly from agricultural soils. Partially decomposed or undecomposed organic matter present in urban soils decreased sorption and increased desorption of 2,4-D. Most of the selected soils contained similar minerals, and quartz was found to be the dominant component that influenced the sorption process. The lower Kd values observed in the urban soils clearly indicate the greater mobility of 2,4-D to the surface and groundwater. Parameters of desorption studies indicated that 2,4-D desorption is more in the selected urban soils. In addition, the calculated values for the HI, GUS, and LIX parameters for all urban soils suggest that 2,4-D is highly mobile and more prone to leaching from surface soil to groundwater, which may affect the drinking water quality, non-target organisms, as well as food safety. Also, 2,4-D from impermeable surfaces can be considered as a direct source of surface and groundwater contamination. Thus, 2,4-D application, particularly on impervious surfaces may require caution because higher concentrations are likely to pose a potential threat to the health of the biota of ecological significance. Further studies are needed involving a greater number of urban soils to arrive at a generalization on the behavior of 2,4-D.

Experimental Section

Chemicals

Analytical grade 2,4-D (98% purity) was obtained from Sigma-Aldrich. All reagents used for the analysis of 2,4-D were of analytical or liquid chromatography–mass spectrometry (LC–MS) grade. A stock solution of 100 mg L–1 was prepared by dissolving 2,4-D in aqueous methanol (1%) and stored at 4 °C under darkness. The stock solution of 2,4-D was diluted with an aqueous solution of CaCl2 (10 mM) to obtain the working concentrations of 0.5, 1, 5, 10, and 20 mg L–1.

Soil Samples and Analysis

Nine urban soils, originally dedicated to farming of garden vegetables, flowers and ornamental plants, lawn grass, and so forth. from locations in and around Newcastle region of Australia (from 32°55′S to 151°45′E) were collected from the surface to 15 cm. A composite sample of each soil was obtained from five randomly collected samples from a location. The soils with assigned IDs (Table ) were air-dried for 24 h, passed through a 2 mm sieve, and stored in hermetic bottles at room temperature until use.[17]

The texture of the soils was determined according to hydrometer method.[46] Soil pH was measured in a pH meter (Laqua, Horiba Scientific), and CEC was determined using the BaCl2 compulsive exchange method.[47] TOC, nitrogen (N), and sulfur (S) in soils were determined through a LECO analyzer (LECO Corporation, USA). Portions of soil samples (0.5 g) were digested in a Microwave Digestion System (MARS 6 240/50, USA), and elements were extracted using 5 mL aqua regia for determination of Fe, Al, Na, K, Ca, and Mg by inductively coupled plasma optical emission spectrometry (ICP–OES, Model: Avio 200, PerkinElmer Pvt Ltd, Singapore). The mineral content of all soils was determined by an X-ray diffractometer (PANalytical BV, The Netherlands) and identified (Table ). Triplicates of each soil sample were used for determination of the above physicochemical characteristics.

Batch Sorption–Desorption Experiments

For sorption–desorption kinetics of 2,4-D, duplicates of air-dried soil samples (2 g), contained in 50 mL polypropylene centrifuge tubes, were equilibrated at 21 ± 1 °C with 10 mL aqueous solution of 2,4-D at 10 mg L–1 in 10 mM CaCl2. CaCl2 solution (10 mM) was used as the background electrolyte to maintain constant ionic strength of sorption and desorption equilibration solutions.[19] The soil samples treated with or without 2,4-D were shaken for 5, 10, 15, 30, 45, 60,180, 360, 720, and 1440 min. Soil suspensions were centrifuged at 5000 rpm for 15 min, and the supernatants were passed through 0.2 μm PTFE filters. The amount of 2,4-D remaining in the soil solutions was determined using LC–MS. For studying desorption kinetics, when sorption reached equilibrium after 24 h, the supernatant was replaced with 10 mM CaCl2 solution in milliQ water. The experimental procedures used for desorption studies were like those of the sorption kinetics. Sorption isotherms of 2,4-D were determined in triplicates of 2 g soil samples contained in 50 mL centrifuge tubes with varying concentrations of 0.5, 1, 5, 10, and 20 mg L–1 in 10 mM CaCl2. Soil samples were equilibrated at 21 ± 1 °C with shaking for 24 h, then centrifuged, filtered, and the supernatant was analyzed for 2,4-D. After 24 h of sorption isotherm study, the supernatant was replaced with 10 mM CaCl2 aqueous solution and desorption isotherms were obtained by following the similar procedure as described above for sorption isotherms.

Liquid Chromatography–Mass Spectrometry

2,4-D concentrations in all liquid samples were determined using a LC–MS system (Agilent 1260/6150B, Agilent Technologies, USA) equipped with a Zorbax Eclipse plus C18 column (4.6 × 150 mm and 3.5 μm dia, Agilent Technologies, USA). The oven temperature was maintained at 35 °C throughout the analysis. The mobile phases were 10 mM aqueous ammonium acetate solution (A) and methanol (B). The mobile phase gradient started at 70% B at 0.0 → 3.0 min, and went to 95% B at 3.0 → 6.0 min and then 70% B at 6.0 min following post-run time of 5.0 min with the flow rate set at 0.5 mL min–1. The analysis was done in the negative mode with the single quadrupole mass spectrometer and other set parameters were: drying gas flow 12.0 mL min–1 at 300 °C, nebulizer pressure 35 psi, capillary voltage 4000 V, SIM ion 219 → 221 amu, fragmentor voltage 100 V, and sheath gas flow 3.0 mL min–1 at 150 °C. The data obtained were processed using Agilent OpenLAB CDS ChemStation software. The standard curve was linear over the tested concentration range (R2, 0.998). The LOD and LOQ values were 0.007 and 0.03 mg L–1, respectively. The recovery (mean, n = 3) of spiked 2,4-D for 0.031–0.5 mg L–1 ranged between 87.98 and 117.94%. Therefore, the method appears to be reliable and accurate for the determination of the 2,4-D in different urban soils.

Data Analysis

Langmuir and Freundlich isotherms and kinetic models were applied to the data obtained in order to understand the sorption–desorption behavior of 2,4-D in soils. Sorption capacity, expressed as the quantity of sorbed herbicide per unit mass of soil, was calculated by the equationwhere qe, Ci, and Ce (mg L–1) are the sorption capacity, initial and equilibrium pesticide concentrations, respectively, and m represents the soil amount (g) and V is the solution volume (L). Linear isotherm can be explained by the following equation, where Kd (L kg–1) denotes distribution coefficient at a defined concentration from the isotherm range.[19]

Langmuir isotherm is a model for single-layer sorption, widely used in sorption studies. The following linear equation was used to define this isotherm.[19,48]where KL (L kg–1) and qmax (mg Kg–1) represent the Langmuir equilibrium constant and sorption capacity, respectively. Freundlich isotherm is another well-known empirical equation used to model heterogeneous sorption that assumes the presence of multilayer sorption sites,[19,49] which is as follows.where KF (mg1–1/n L1/n g–1) indicates the Freundlich constant used as a measure of sorption efficiency and 1/n is the sorption intensity that indicates the favorability of sorption. These isotherm parameters can be obtained from the intercept and slope of the Freundlich graphs plotted for log qe versus log Ce (mg L–1). Pseudo-first- and pseudo-second-order reaction kinetic models were used to obtain information on the rate of uptake and release of 2,4-D during sorption–desorption mechanism.[19,50] Pseudo-first-order reaction kinetic can be defined by the following linear equationwhere, k1 (min–1) is the rate constant of pseudo-first-order kinetic, qe and q (mg Kg–1) are the amounts of sorbed pesticide at saturation and time “t” (min), respectively. Pseudo-second-order reaction kinetic can be expressed by the following equation

Leaching potential was determined by the GUS based on vii,[41] and LIX using viii,[51] both of which are dependent on the herbicide’s half-life (days) in soil (t1/2) and Koc values.[41]where Koc (L kg–1) is soil OC content and k is the herbicide first-order rate constant (day–1) given by the ix and x. Kd values were calculated for each soil-herbicide interaction and normalized to the soil OC content (%) to obtain the Koc coefficient[43] as follows

The HI was calculated for the sorption/desorption isotherms according to xi(52)where 1/n is the Freundlich exponent for sorption and desorption isotherms.[45]