JEQ Journal of Natural Resources and Life Sciences Education
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online 11 May 2005
Published in J Environ Qual 34:1045-1054 (2005)
DOI: 10.2134/jeq2003.0423
© 2005 American Society of Agronomy, Crop Science Society of America, and Soil Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (3)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Ahmad, R.
Right arrow Articles by Kookana, R. S.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Ahmad, R.
Right arrow Articles by Kookana, R. S.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by Ahmad, R.
Right arrow Articles by Kookana, R. S.
Related Collections
Right arrow Laboratory Column Studies
Right arrow Pesticides
Right arrow Sorption/Exchange
Right arrow Organic Compounds
Right arrow Agricultural Pesticides

TECHNICAL REPORTS

Organic Compounds in the Environment

Measuring Sorption of Hydrophilic Organic Compounds in Soils by an Unsaturated Transient Flow Method

Riaz Ahmada,b, Hidetaka Katouc and Rai S. Kookanad,*

a Department of Soil and Water, University of Adelaide, PMB1, Glen Osmond, SA 5064, Australia
b Current address: Faculty of Horticulture, Chiba University, 648 Matsudo, Matsudo, Chiba, 271-8510 Japan
c National Institute for Agro-Environmental Sciences, Kannondai 3-1-3, Tsukuba, 305-8604 Japan
d CSIRO Land and Water, PMB2, Glen Osmond, SA 5064 Australia

* Corresponding author (Rai.Kookana{at}csiro.au)

Received for publication December 3, 2003.

   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 THEORY
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 REFERENCES
 
Determination of sorption of hydrophilic, weakly sorbing organic compounds in soil by conventional batch methods using a slurried suspension is often prone to considerable errors because small changes in the solution concentration on equilibration must be accurately determined. This difficulty is exacerbated for compounds susceptible to degradation, which also decreases the solution concentration. The objective of this study was to determine sorption of hydrophilic pesticides by applying an unsaturated transient flow method, which enables determination of sorption at sufficiently small solution to soil ratios. The method makes use of piston-like displacement of the antecedent solution in equilibrium with sorbed phase when pesticide-free water is infiltrated into a soil column spiked with a pesticide. Pesticide sorption and the solution concentration are inferred from a plot of total pesticide content per unit mass of soil vs. water content in a region where the antecedent solution is accumulated. Thus, extraction of solution from relative dry soil is unnecessary. We tested this method for two hydrophilic pesticides, monocrotophos [dimethyl (E)-1-methyl-2-(methyl-carbamoyl) vinyl phosphate] and dichlorvos (2,2-dichlorovinyl dimethyl phosphate). The sorption coefficient, Kd, obtained for monocrotophos was slightly lower than that by batch method (Kd = 0.10 vs. 0.19 L kg–1), whereas for dichlorvos, a compound highly susceptible to degradation, the unsaturated flow method yielded a much smaller Kd (0.19 vs. 3.22 L kg–1). The Kd values for both compounds were consistent with the observed retardation in the pesticide displacement in the columns. The proposed method is more representative of field conditions and particularly suitable for weakly sorbing organic compounds in soils.

Abbreviations: HPLC, high performance liquid chromatography


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
THEORY
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 REFERENCES
 
HYDROPHILIC ORGANIC SUBSTANCES showing a low affinity to sorption sites in soil are susceptible to leaching and pose a greater risk of ground water pollution than strongly sorbing hydrophobic substances. Determination of sorption of weakly sorbing organic compounds is, however, not always easy and often prone to considerable experimental errors.

The most commonly used method for measuring sorption of organic compounds in soil is the batch sorption technique in which a slurry or suspension of soil is equilibrated with a solution at different initial concentrations of the compound of interest under constant agitation (Green and Karickhoff, 1990a; Cleveland, 1996; USEPA, 1996; Wauchope et al., 2002). In this technique, sorption by soil is calculated from the difference between the initial and final solution concentrations of the compounds (Wauchope et al., 2002). For an accurate measurement of sorption, the decline in the solution concentration following equilibration needs to be substantial [e.g., in the range of 40 to 60% (Organisation for Economic Co-operation and Development, 2000), or the sorption between 20 to 50% (Wauchope et al., 2002)]. The sorption coefficient Kd (= sorption per unit mass of soil, Q, divided by the solution concentration, C) needs to be larger than 5 L kg–1 to achieve >50% decrease in the solution concentration on equilibration in a batch system with a soil-to-solution ratio of 1:5. However, for hydrophilic, weakly sorbing organic compounds, the Kd values are often smaller. For example, atrazine, a commonly detected herbicide in ground water, has an average Koc of 100 (Hornsby et al., 1996), which means in a soil with organic carbon content of 10 g kg–1 the Kd will be equal to 1. Many compounds show even lower sorption in soils (e.g., acidic herbicides and pesticide metabolites). For these compounds, the concentration changes before and after equilibration in the conventional batch sorption experiments are often much smaller than those recommended. As a result, considerable errors in the estimated sorption could arise due to experimental errors in pesticide measurements (Boesten, 1990; Brownawell et al., 1990; Johnson and Farmer, 1993). To avoid such errors in the estimated sorption, a sufficiently small solution-to-soil ratio is required (e.g., a soil-to-solution ratio of 1 for Kd = 1), but this is difficult to realize for weakly sorbing compounds having a much smaller Kd.

This difficulty with the conventional batch methods in measuring sorption of weakly sorbing compounds is even more exacerbated when the compounds are susceptible to degradation. Either biotic or abiotic, decreases in the solution concentration due to degradation may seriously affect experimental results, and, if not effectively controlled, can lead to erroneously high estimates of sorption (Dekkers, 1978; Koskinen and Cheng, 1983; Green and Karickhoff, 1990b; Wauchope et al., 2002). Biocides such as sodium azide (NaN3) may be used to suppress biotic degradation and/or transformation during the equilibration, but their use can alter soil properties (e.g., soil pH) and does not always guarantee the sterility of the system. Abiotic losses such as hydrolysis are much more difficult to control, particularly if the rate of hydrolysis is different between the solution phase and the sorbed phase. Consequently, no solution for avoiding errors stemming from degradation and/or losses has been proven ideal in the batch sorption measurements (Koskinen and Cheng, 1981).

Another technique that has also been employed for measuring sorption is the flow-based, steady-state miscible displacement methods (Kookana et al., 1992; Rheinländer et al., 2000). These methods have some advantages over the batch methods in that the experimental conditions are more representative of field. However, in these methods, equilibration is often slow relative to residence time of the solution in the columns, and the kinetic effects may dominate so that there remains uncertainty whether sorption equilibrium has been attained (Skopp, 1986; Bürglsser et al., 1993). Thus, a novel, alternative method needs to be developed that can sensitively measure sorption of hydrophilic, weakly sorbing compounds in soil at a sufficiently small solution-to-soil ratio, while it is exempt from uncertainties about attainment of sorption equilibrium.

The unsaturated transient flow method is a technique developed by Katou et al. (2001) for determining adsorption of weakly sorbing inorganic ions in soil. This method is essentially a variant of batch methods in that sorption equilibrium is allowed to be established in the absence of water flow after spiking soil with a solute of interest. Subsequent transient water flow is introduced only to allow the equilibrated solution to be accumulated in a region beyond the "plane of separation" (Smiles and Philip, 1978) in the column. Instead of extracting solution from relatively dry soil and measuring the solution concentration, solute content per unit mass of soil, including solute present in the aqueous solution phase as well as in the sorbed phase, at different water contents in the region beyond the plane of separation is determined after extraction using an appropriate solvent. Solute sorption by soil and the equilibrated solution concentration before the water flow are inferred from a plot of the solute content per unit mass of soil against the water content in this region. This enables sorption equilibrium to be established at a soil-to-solution ratio sufficiently low for accurate determination of sorption for solutes having a low affinity to soils.

The objective of the present study was to explore the application of the unsaturated transient flow method for determining sorption of organic compounds that are hydrophilic, weakly sorbing, and/or susceptible to degradation during equilibration, for which conventional batch methods are fraught with difficulties and prone to errors. Sorption of two hydrophilic, weakly sorbing, and relatively fast-degrading pesticides, monocrotophos and dichlorvos, in soil was measured by the unsaturated transient flow method under realistic soil water conditions. The results were compared with those obtained by a conventional batch technique using slurried suspensions.


   THEORY
TOP
 ABSTRACT
 INTRODUCTION
 THEORY
MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 REFERENCES
 
The unsaturated transient flow method (Katou et al., 2001) makes use of the piston-like displacement of antecedent water by the invading water during infiltration into unsaturated soil. It has been well established that during absorption of water into an unsaturated homogeneous soil having small aggregate size (e.g., <2 mm), the antecedent water is pushed ahead by the invading water, and accumulates in a region beyond the "plane of separation," x* (Smiles and Philip, 1978; Clothier et al., 1988; Bond and Phillips, 1990). This plane, which identifies the front of the invading water, is given by the relation (Smiles and Philip, 1978):

[1]
where x is the distance (m), {theta} is the volumetric water content (m3 m–3), and {theta}n and {theta}s denote the initial water content and the water content at the inlet end of the soil column, respectively (Fig. 1) .



View larger version (24K):
[in this window]
[in a new window]
  		Fig. 1. Schematic diagram of determination of solute sorption by the unsaturated transient flow method (after Katou et al., 2001). Initial liquid-phase concentration, Cn, and initial sorption by soil, Qn, are obtained from the M vs. ({theta}/{rho}) plot for x > x*. {theta}n = initial water content; {theta}s = water content at the column inlet; Mn = initial solute content in soil; {rho} = bulk density; x* = plane of separation.

 
If a sorption equilibrium has been allowed to be established in soil before the infiltration of water, by spiking the soil with a solute of interest, this accumulation of the antecedent water in the region x > x* allows the equilibrium solution concentration and solute sorption in soil to be estimated from the solute content profile in this region. Since the aqueous solution found in this region is derived from the antecedent solution pushed ahead, it has the same concentration and composition as the initial soil solution, which has been equilibrated with the sorbed phase before the water flow. With the solution concentration and composition remaining constant, the solute sorption in soil should also be constant for x > x*. Consequently, any changes in the solute content per unit mass of soil (i.e., the sum of those present in the aqueous solution phase and those in the sorbed phase) in this region are simply due to changes in the water content caused by the accumulation of the antecedent solution. Thus, a plot of the solute content per unit mass of soil, M (mg kg–1), vs. water content (more exactly, the solution volume per unit mass of soil, {theta}/{rho}) in the region beyond the "plane of separation" yields a linear relationship (Fig. 1):

[2]
in which the initial solution concentration, Cn (mg m–3), and the initial solute sorption, Qn (mg kg–1), before the water infiltration are given as the slope and the intercept of the plot, respectively, for a soil at a bulk density {rho} (kg m–3). Spiking soil with different amounts of solute will give different values of initial solute content Mn (mg kg–1) (and hence Cn and Qn), and will enable sorption equilibria to be established over a wide range of concentrations. In this way, if desired, sorption isotherms of an arbitrary form may be constructed from a series of column experiments.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 THEORY
 MATERIALS AND METHODS
RESULTS AND DISCUSSION
 REFERENCES
 
Pesticides and Their Chemical Properties
Two pesticides, monocrotophos (molecular weight 223.2) and dichlorvos (molecular weight 210.0), were used in this study. Both chemicals belong to the organophosphorus group of pesticides commonly used in agriculture. According to Tomlin (2002), both compounds are highly soluble. While the aqueous solubility of monocrotophos (completely miscible) is greater than that of dichlorvos (18 g L–1 at 25°C), the vapor pressure of dichlorvos is much higher (2.1 x 103 mPa at 25°C) than that of monocrotophos [2.9 x 10–1 mPa (20°C)]. Dichlorvos is susceptible to losses due to rapid degradation, transformation, and volatilization (Tomlin, 2002). Technical grade chemicals (>98% purity) were used in this study.

Horizontal Water Infiltration Experiments
One-dimensional, horizontal, constant-flux water infiltration experiments were performed in sectionable columns to determine sorption of the two pesticides. The acrylic columns, called Perroux tube (Clothier and Scotter, 2002), were 20 cm long and had an internal diameter of 1.91 cm. Each slab forming the column had a thickness of 0.8 cm. These slabs were arranged to form a column in such a way that enabled instantaneous sectioning at the end of experiments. A schematic illustration of the apparatus has been provided by Clothier and Scotter (2002).

The soil used in the experiments was taken from the 0- to 5-cm layer of a Palexeralf from Pinnaroo in South Australia. The soil was air-dried and sieved to 2 mm. Some pertinent properties of the soil are given in Table 1. The soil was spiked with an aqueous pesticide solution to give an initial pesticide content of either 2.5 mg monocrotophos kg–1 soil or 6 mg dichlorvos kg–1 soil, and an initial soil water content of approximately 0.07 kg kg–1. The solution was sprayed in fine droplets over a thin layer of the soil. Sorption of monocrotophos was also studied at a higher initial soil water content of approximately 0.12 kg kg–1. The spiked soil was mixed thoroughly in a sealed polyethylene bag to obtain uniform distribution of the pesticide. After allowing the contact time of 16 h for monocrotophos and 4 h for dichlorvos for equilibration between the aqueous solution phase and the sorbed phase, the soil was packed uniformly into the sectionable acrylic column. A shorter contact time was allowed for dichlorvos, as the chemical undergoes rapid hydrolysis. A summary of the experimental conditions is given in Table 2.


View this table:
[in this window]
[in a new window]
  		Table 1. Some pertinent physical and chemical properties of the Pinnaroo soil.

 

View this table:
[in this window]
[in a new window]
  		Table 2. Summary of the one-dimensional, horizontal infiltration experiments.

 
The column experiments were conducted in laboratory at 22 ± 2°C by feeding pesticide-free solution (Milli-Q water; Millipore, Billerica, MA) into the columns packed with the spiked soil at a constant flux of 1.16 x 10–5 or 5.82 x 10–6 m s–1 using a peristaltic pump (Minipuls 3, Gilson; John Morris Scientific Pty. Ltd., Australia). After the wetting front had reached about a 12-cm distance from the proximal end of the column, the flow was terminated and the elapsed time was noted. Following quick sectioning of the column at the end of experiment, a portion of soil from each section (approximately 1.5 g) was immediately transferred to a preweighed glass tube containing 3 mL of CH3OH·H2O (90:10) and weight of the tube containing the extracting solution plus soil was recorded. Monocrotophos and dichlorvos in the samples were extracted by shaking the tube for one hour. The samples were then centrifuged at 1800 rpm for 10 min, syringe-filtered, and analyzed for pesticide content in soil, including those in the aqueous solution phase and in the sorbed phase, following the high performance liquid chromatography (HPLC) methods described elsewhere (Ahmad et al., 2001a). The remaining soil from each 8-mm section of the column was transferred into a preweighed container and the water content measured gravimetrically.

The pesticide content in the soil samples taken from the region beyond the plane of separation, where the antecedent water accumulated, was plotted against the solution volume per unit mass of soil ({theta}/{rho}). The initial aqueous-phase concentration, Cn, and the initial sorption by soil, Qn, were deduced from linear regression analysis for the pesticides (Eq. [2]). The sorption coefficient, Kd, of each pesticide was obtained from the quotient Qn/Cn.

Pesticide Sorption to Column Material
To determine if the pesticides were sorbed onto the column material, slabs of the column were immersed in aqueous solutions of dichlorvos at an initial concentration of 6.0 mg L–1 for 4 h and monocrotophos at an initial concentration of 2.5 mg L–1 for 16 h. At the end of immersion, the concentrations of each pesticide were determined by HPLC following the methods referred to above and compared with the concentrations in control solutions.

Batch Experiments
Batch studies with monocrotophos and dichlorvos in the Pinnaroo soil were also performed to compare the Kd values obtained from the batch and transient flow methods. Five milliliters of an aqueous solution containing 0.25, 0.5, 1.0, 2.0, and 3.0 mg L–1 of monocrotophos and 6 mg L–1 of dichlorvos (in triplicate) were added to 5-g portions of soil in glass tubes with Teflon caps. A solution-to-soil ratio of 1:1, much smaller than the commonly adopted ratios of 5:1 or 10:1, was deliberately chosen according to the OECD recommendations (Organisation for Economic Co-operation and Development, 2000) because small values of Kd were expected for the two hydrophilic pesticides. The soil suspensions were shaken for 16 h for monocrotophos and 1 h for dichlorvos, and centrifuged at 1800 rpm for 20 min to obtain a clear solution. The loss of dichlorvos in the batch system under constant agitation was much more rapid than that in the unsaturated soil. This rapid loss prohibited a shaking time comparable to the contact time of 4 h adopted in the unsaturated transient flow experiment. According to sorption studies on weakly sorbing compounds (Ahmad et al., 2001b), sorption equilibration was likely to reach about 80 to 90% during the shaking time of 1 h. An aliquot of the clear supernatant was syringe-filtered through regenerated cellulose (RC) membrane filters (Ahmad et al., 2001a) and analyzed by HPLC for the pesticides. The Kd value of monocrotophos was estimated from the sorption isotherm constructed over a range of the pesticide concentrations, whereas the Kd value of dichlorvos was determined from a single concentration. No attempt was made to use biocides to suppress losses due to biotic degradation during the batch experiments.


   RESULTS AND DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 THEORY
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
REFERENCES
 
Sorption of Monocrotophos and Dichlorvos onto the Column Material
The tests to determine if there was sorption of the pesticides to the column material (acrylic) during the experiment showed that there was no significant sorption loss of dichlorvos (<1.0%) or monocrotophos (<0.5%) to the column material, as compared with the control. The studies by Topp and Smith (1992) on sorption of atrazine and metolachlor (more hydrophobic than the compounds used in our study) by a variety of tubing materials also support the suitability of acrylic materials for this study. They found that neither of these pesticides was significantly sorbed by acrylic plastic.

Sorption at Low Initial Water Content by the Unsaturated Flow Method
Figure 2 shows the water content profiles on one-dimensional infiltration of water into soil columns spiked with monocrotophos and dichlorvos. The gravimetric water contents of the soil after spiking the monocrotophos and dichlorvos were 0.075 and 0.065 kg kg–1, respectively. In the figure, the initial volumetric water contents {theta}n = 0.103 and 0.105 m3 m–3, respectively, denote the water content of the spiked soils when packed into the columns. The planes of separation x*, calculated from Eq. [1], were found to be 6.8 and 7.5 cm for monocrotophos and dichlorvos, respectively.



View larger version (21K):
[in this window]
[in a new window]
  		Fig. 2. Water content profiles on one-dimensional infiltration of water into soil columns spiked with monocrotophos and dichlorvos. x* = plane of separation; t = time; {theta}n = initial water content.

 
The measured pesticide content profiles following the infiltration of water are presented in Fig. 3 for the columns spiked with monocrotophos and dichlorvos at low initial soil water contents. As mentioned before, either 2.5 mg kg–1 soil of monocrotophos or 6 mg kg–1 soil of dichlorvos had been added to the soils before packing into columns. Upon imbibition of water, monocrotophos was almost completely removed from the soil near the column inlet whereas for dichlorvos a small amount (approximately 0.09 mg kg–1) was detected. This is consistent with the solubility of the two compounds, dichlorvos being more hydrophobic than monocrotophos. We also see that both for monocrotophos and dichlorvos, the front of pesticide being removed lagged behind the plane of separation x*, the location of the displacement front expected for an inert solute. This is indicative of some retardation of the pesticide transport due to sorption.



View larger version (23K):
[in this window]
[in a new window]
  		Fig. 3. Monocrotophos and dichlorvos content profiles on one-dimensional infiltration of water into soil columns spiked with pesticides. M = pesticide content in soil; Mn = initial pesticide content; Qn = initial sorption by soil; {theta}n = initial water content; x* = plane of separation.

 
In Fig. 4 , the pesticide content in soil, M, in the region beyond the plane of separation (x > x*) was plotted against the water content ({theta}/{rho}) for the columns premixed with monocrotophos and dichlorvos. In accordance with the theory (Katou et al., 2001), a linear relationship was observed between M and ({theta}/{rho}) for each pesticide, except for "dry" samples at {theta} {approx} {theta}n taken near the dry end of the dichlorvos column. This demonstrates that the accumulation of the pesticides in this region was due solely to the accumulation of the antecedent solution pushed ahead, with the pesticide concentration in the solution and sorption in soil remaining constant. Thus, the intercept and the slope of the M vs. ({theta}/{rho}) plot represent the pesticide sorption in soil and the concentration in the solution, respectively. Linear regression analyses performed on the pesticide and water content data in this region (excluding five "dry" samples at {theta} {approx} {theta}n for dichlorvos) yielded estimates (±SE) of Qn = 1.69 (±0.08) mg kg–1 and Cn = 15.6 (±0.6) mg L–1 for monocrotophos (r2 = 0.985), and Qn = 0.47 (±0.04) mg kg–1 and Cn = 2.46 (±0.25) mg L–1 for dichlorvos (r2 = 0.970). Thus, the sorption coefficients Kd (= Qn/Cn) were 0.109 (±0.010) L kg–1 for monocrotophos and 0.19 (±0.04) L kg–1 for dichlorvos. A repeat of the experiment with monocrotophos conducted under similar experimental conditions ({theta}n = 0.096 m3 m–3) yielded reproducible results, with a Kd value of 0.099 (±0.018) L kg–1 derived from Qn = 1.38 (±0.15) mg kg–1 and Cn = 13.9 (±1.0) mg L–1. The sorption coefficients (Kd) determined above were checked against the independently measured retardation factors obtained from the pesticide content profiles in the region behind the plane of separation (0 < x < x*) in the column. This is discussed in detail in the latter section of the paper.



View larger version (21K):
[in this window]
[in a new window]
  		Fig. 4. Relationships between pesticide content in soil, M, and the solution volume per unit mass of soil, ({theta}/{rho}), in the region x > x* for monocrotophos and dichlorvos. Note that for dichlorvos, data on samples taken from the dry end of the column (at {theta} {approx} {theta}n) were excluded from the regression analysis. Cn = initial liquid-phase concentration; Qn = initial sorption by soil; {theta}n = initial water content; x = distance; x* = plane of separation.

 
The estimates of Cn and Qn were also checked against the mass balance of the pesticide recovered from the column. From these estimates, the initial pesticide content in soil, Mn (mg kg–1), is calculated using the relation:

[3]

Ideally, Mn should be equal to the average pesticide content in the column, Mavg (mg kg–1), obtained as:

[4]
where L is the column length (m). In Fig. 3, Mn inferred using Eq. [3] was also shown for each column. For the monocrotophos column with {theta}n = 0.103 m3 m–3, Mn = 2.87 (±0.10) mg kg–1 and Mavg = 2.84 mg kg–1. Excellent agreement between the two values suggests that Cn and Qn were correctly estimated. In addition, these values were comparable to the amount added to the soil (= 2.5 mg kg–1), indicating that the loss of monocrotophos during the experiment was minimal. Contrastingly, we obtained Mn = 0.63 (±0.01) mg kg–1 for dichlorvos, which was merely approximately 10% of the pesticide incorporated per unit mass of soil (Table 2). From the measured M(x) profile, we found Mavg = 0.60 mg kg–1, which was in close agreement with Mn inferred above. These results show that there was a substantial loss (approximately 90% of the amount added to soil) of dichlorvos during the experiment.

As mentioned earlier, dichlorvos is highly susceptible to losses due to degradation and other processes, and as such was a difficult compound to study. In the dichlorvos column, the pesticide contents near the dry end of the column (i.e., at {theta} {approx} {theta}n) were considerably smaller than Mavg obtained above (Fig. 3). The reason for these small dichlorvos contents in the dry samples is unknown. When these data were included in the regression analysis, the estimates of Cn = 4.51 (±0.28) mg L–1 and Qn = 0.12 (±0.04) mg kg–1 were obtained. These estimates give an initial pesticide concentration of Mn = 0.42 (±0.05) mg kg–1, which is considerably smaller than the experimentally obtained Mavg, and a sorption coefficient of Kd = 0.027 (±0.009) L kg–1, which is much smaller than Kd = 0.19 (±0.04) L kg–1 based on the regression excluding the dry samples. The estimates based on the regression including data from the dry samples were inconsistent not only with Mavg, but also with a slightly larger retardation for dichlorvos than for monocrotophos observed in the column experiments. Estimates of Cn and Qn deduced from the unsaturated transient flow method are sensitive to solute contents near the dry end of the columns. Therefore, particular attention should be taken to minimize experimental errors in extracting the solutes and determining the contents in these samples.

Sorption at Higher Initial Water Content
To examine the influence of soil water content during equilibration on the estimates of sorption for monocrotophos by the unsaturated transient flow method, an experiment was conducted at a higher initial volumetric water content of {theta}n = 0.190 m3 m–3 (initial gravimetric water content wn = 0.116 kg kg–1). The relationship between the pesticide content in soil and the water content for the region x > x* in the column is shown in Fig. 5 . From the estimates Qn = 1.07 (±0.18) mg kg–1 and Cn = 11.3 (±1.0) mg L–1, Kd was calculated at 0.095 (±0.024) L kg–1, which was only slightly lower than that obtained for a lower water content ({theta}n {approx} 0.10 m3 m–3). We interpret these results as demonstrating that the unsaturated transient flow method can be used to determine sorption coefficients at different, but still realistically low soil water contents. This is an interesting feature of the method because it has been reported that soil water content can significantly affect the sorption of some pesticides in soils (Berglöf et al., 2000).



View larger version (17K):
[in this window]
[in a new window]
  		Fig. 5. Relationship between pesticide content in soil, M, and the solution volume per unit mass of soil, ({theta}/{rho}), in the region x > x* for monocrotophos at a higher initial soil water content ({theta} = 0.19 m3 m–3). Cn = initial liquid-phase concentration; Qn = initial sorption by soil; {theta}n = initial water content; x = distance; x* = plane of separation.

 
Comparison of Unsaturated Transient Flow and Batch Sorption Coefficients
The batch sorption experiments conducted with aqueous solutions containing 0.25, 0.5, 1.0, 2.0, and 3.0 mg L–1 of monocrotophos gave the final aqueous-phase concentrations of 0.16, 0.33, 0.72, 1.51, and 2.35 mg L–1, respectively. The corresponding pesticide sorption values by soil (calculated from the difference between the initial and final aqueous-phase concentrations) were, respectively, 0.09, 0.17, 0.28, 0.49, and 0.65 mg kg–1. The sorption isotherm obtained from the batch method was nonlinear, and was described well by the following Freundlich equation:

[5]
where Q is the pesticide sorption by soil (mg kg–1) and C is the pesticide concentration in the solution (mg L–1). Thus, the sorption coefficient Kd was weakly concentration-dependent and was in the range of 0.52 to 0.28 L kg–1 for aqueous-phase concentrations of 0.16 to 2.35 mg L–1. Using Eq. [5], Kd for higher concentrations as observed in the unsaturated transient flow experiments (C = 11.3–15.6 mg L–1) was extrapolated and estimated at 0.196 to 0.181 L kg–1. These values were twice as large as those obtained with the transient flow method (Table 3).


View this table:
[in this window]
[in a new window]
  		Table 3. Comparison of sorption coefficient (Kd) values obtained from the unsaturated flow method and the batch method.

 
For dichlorvos, the batch experiments with the initial aqueous-phase concentration of 6.0 mg L–1 gave the final concentration of 1.42 mg L–1 on equilibration, and by difference, the pesticide sorption by soil was calculated at 4.58 mg kg–1. Similar experiments conducted using dichlorvos solutions at lower initial concentrations did not yield reproducible results, owing to rapid loss of the pesticide. Thus, the batch Kd for dichlorvos was determined at the single concentration and found to be 3.22 L kg–1 (Table 3). This value was 17 times greater than that obtained at a comparable aqueous-phase concentration with the unsaturated flow method (Kd = 0.19 L kg–1 at C = 2.46 mg L–1).

The larger Kd value for dichlorvos obtained by the batch method was mainly due to greater losses of the pesticide during the experiment, which was erroneously attributed to sorption. The losses could result from such processes as biotic or abiotic degradation and/or transformation, volatilization, etc. In the present study, we did not attempt to minimize biotic losses of the pesticide by the use of biocides. Use of biocides could lead to a smaller decrease in the solution concentration during the batch experiments, and hence a Kd value more comparable with that obtained from the unsaturated transient flow method. However, satisfactory control experiments are not always feasible for a class of rapidly degradable and/or hydrolyzable compounds. The losses in the control experiments without soil are entirely different from those in the soil slurry, owing to lower biological losses and more rapid hydrolysis in the control experiments. This at least partly explains why laboratory-determined Kd values for such compounds are scarce in the literature. It should be noted that the Kd values for degradable and/or hydrolyzable compounds reported in the literature (Hornsby et al., 1996) are only estimated values. This reflects difficulties in carrying out batch studies on such compounds. In the unsaturated transient flow method, estimation of sorption is based on measurements of total pesticide content in the soil (=including those in the aqueous solution and in the sorbed phase). Since these pesticide contents in soil have already reflected any possible losses or degradation during the experiments, the method is less prone to overestimation of sorption due to degradation, even when a biocide is not used to suppress degradation of the pesticides.

Comparison of Sorption Coefficient with Retardation of Pesticide Displacement Observed in the Columns
An advantage of the unsaturated transient flow method is that the sorption coefficient Kd deduced from the water and pesticide content profiles in the region beyond the plane of separation (x > x*) may be tested for its ability to predict the retarded solute displacement front x' in the region behind this plane (0 < x < x*). It is interesting to compare the retardation factor R and x' based on Kd obtained by the batch and the unsaturated flow methods, with the retardation of pesticide transport observed independently in the column experiments. For the dichlorvos column with {theta} {approx} 0.35 m3 m–3 behind x* and {rho} = 1.60 x 103 kg m–3, the unsaturated-flow Kd of 1.90 x 10–4 m3 kg–1 gives R = 1 + ({rho}/{theta}) Kd = 1.87, and predicts x' = x*/R = 4.03 cm. The experimental displacement front for dichlorvos was located at x' {approx} 4.1 cm (Fig. 3), showing an excellent agreement with the predicted solute front. In contrast, the batch Kd of 3.22 x 10–3 m3 kg–1 gives much poorer predictions of R = 15.7 and x' = 0.48 cm. Obviously, the batch Kd greatly underestimated the distance of dichlorvos transport observed in the column. Similarly, for the monocrotophos column with {theta}n = 0.103 m3 m–3, the unsaturated-flow Kd of 1.09 x 10–4 m3 kg–1, together with {theta} {approx} 0.305 m3 m–3 and {rho} = 1.37 x 103 kg m–3, predicts R = 1.49 and x' = 4.56 cm. This is in close agreement with the observed monocrotophos displacement front (x' {approx} 4.7 cm) in the column. The batch Kd of 1.81 x 10–4 m3 kg–1, calculated using Eq. [5], yielded poorer estimates of R = 1.81 and x' = 3.76 cm, but the predictions were much better than in the case of dichlorvos.

Poor prediction of the dichlorvos displacement front by the batch-determined Kd was again due mainly to greater loss of the pesticide during the experiment. Use of biocides for suppressing biotic degradation could produce a better prediction of x', although this was not tested in the present study. It should be emphasized here that even in the absence of any control of biotic degradation, the unsaturated transient flow method yielded sorption parameters that were capable of predicting the position of dichlorvos displacement front in the column. This is an attractive feature of the method, since satisfactory suppression of both biotic and abiotic degradations is often difficult in the batch slurry experiments, and the resultant overestimation of sorption could lead to a serious underestimation of the risk of ground water contamination by the pesticides.

Advantages and Limitations of the Method
This study has shown that the proposed unsaturated transient flow technique can be successfully used for measuring sorption of hydrophilic pesticides with low sorption affinities for soils. Several organic compounds, including pesticide degradates, fall into this category, and pose a greater risk of leaching and ground water contamination (Kolpin et al., 2000). For a risk assessment of such compounds using suitable solute transport models, realistic and accurate sorption parameters are needed. However, the solutes that have low sorption affinities for soil surface and are susceptible to abiotic and biotic losses are the most difficult to handle in the conventional batch technique. This was the impetus that led us to applying the unsaturated flow method to the determination of sorption of such compounds in the present study. Advantages and possible shortcomings of this method are discussed below.

The unsaturated transient flow method is best suited for measuring sorption of hydrophilic, weakly sorbing substances in soil. For this class of compounds, determination of sorption by the conventional batch methods may be inaccurate if the differences in the solution concentration before and after equilibration are relatively small compared with errors in the measurements (Boesten, 1990; Johnson and Farmer, 1993). To ensure accurate determination of sorption, a sufficiently small solution-to-soil ratio is required, but this will pose another experimental difficulty of extracting the solution from relatively dry soils. In the unsaturated transient flow method, total pesticide content profile in soil is determined to deduce the pesticide sorption and the solution concentration from a plot of pesticide content against water content in a region where the equilibrated solution is accumulated. This enables sorption equilibrium to be established and determined at a small solution-to-soil ratio. As seen in Fig. 4 and 5, the method is sensitive enough to accurately determine the sorption of a hydrophilic pesticide having a Kd of as low as approximately 0.1. In addition, the solution-to-soil ratio during equilibration may be varied so that effects of soil water content on the sorption of organic compounds can be determined. Berglöf et al. (2000) suggested that the effects of soil water content could be significant and should be incorporated in pesticide transport models.

Another advantage of the unsaturated transient flow method is that Kd deduced from the water and pesticide content profiles in the region beyond the plane of separation (x > x*) may be tested for its ability to predict the retarded solute displacement front x' in the region behind this plane (0 < x < x*). In the present study, the retardation factor R based on Kd obtained by the proposed method successfully predicted the location of pesticide displacement fronts x' in the columns. Such comparison cannot be made in the conventional batch experiments alone.

On the other hand, the proposed method is unsuitable for hydrophobic, strongly sorbing organic compounds because the solution concentration after equilibration will be too small to be accurately determined by this method. For such compounds, conventional batch methods at a much larger solution-to-soil ratio should be employed.

Another shortcoming of the proposed method is that it is more time-consuming than the conventional batch methods. One column experiment provides only one set of solution concentration and sorption by soil, and hence a single value of Kd. However, by conducting a series of column experiments in which different amounts of solute have been incorporated into soil, a sorption isotherm of an arbitrary form covering a wide range of solution concentration can be constructed. The inferred isotherm may be then tested for its ability to reproduce the solute content profiles observed in the different columns (Katou et al., 2001; Katou, 2004).

The unsaturated transient flow method makes use of the piston-like displacement of the antecedent water, which has been brought into equilibrium with the sorbed phase, by the invading water in unsaturated soils. Thus, nonuniform water flow in the soils could potentially affect the estimates of the solute sorption and the solution concentration. The presence of mobile and immobile regions has been commonly postulated in modeling water and solute movement in saturated, structured soils (Brusseau and Rao, 1990). However, in unsaturated, homogeneously repacked soils having small aggregate size, the piston-like displacement of the antecedent water is a well-established phenomenon (Smiles and Philip, 1978; Clothier et al., 1988). Even when there is some fraction of immobile water in soil, its presence will not affect the M vs. ({theta}/{rho}) plot as long as the mobile water has the same solute concentration as the immobile water as a consequence of the equilibration before the water flow. Thus, the net effect of the presence of immobile water will be to lower the solution concentration (and hence, the solute sorption as well) as the water content increases from the wetting front toward the plane of separation, x*. The resultant M vs. ({theta}/{rho}) plot will be nonlinear, depending on the degree of intrusion of the invading water, if the fraction of immobile water is considerable. The plots for monocrotophos in Fig. 4 and 5 as well as those reported for inorganic anions (Katou et al., 2001; Katou, 2004) were linear, suggesting that immobile water, if present, had only negligible effects in these experiments.

Another concern about the proposed method is the attainment of sorption equilibrium. In the method, a sufficient contact time between soil and sorbate is allowed to establish sorption equilibrium under no-flow conditions. The method is thus a variant of the batch methods, and in principle, exempt from uncertainty about the attainment of sorption equilibrium as seen in the conventional, flow-based miscible displacement experiments (Kookana et al., 1992). However, even in the batch experiments, rate-limiting slower processes following faster ones are often observed over an extended period of time in the sorption–desorption of organic compounds in soil (Pignatello, 1989). Such long-term sorption kinetics is likely to be associated with rate-limited diffusion of the hydrophobic substances within the matrix of soil organic matter (Brusseau and Rao, 1991). In the proposed method, equilibration at a smaller soil water content in the absence of continuous agitation may lead to a slower rate of approach to sorption equilibrium than in the batch slurry experiments. However, if there has been a considerable degree of sorption non-equilibrium before the water flow, then significant increases in the water content observed in the region x > x* on infiltration of water should trigger a considerable redistribution of the solute between the aqueous phase and the sorbed phase. Consequently, a linear M vs. ({theta}/{rho}) plot indicative of constant solution concentration and solute sorption will not be obtained. Based on the linear relationships observed for monocrotophos and dichlorvos (except in the dry end of the column), we consider that sorption non-equilibrium before the water infiltration was not a significant problem in these columns. This does not necessarily mean, however, the absence of long-term sorption non-equilibrium due to slower processes in the systems. In the unsaturated transient flow method, the contact time may be varied, at a realistic soil water content, to see possible effects of such slower processes not only on the sorption isotherm deduced from the solute content profile in the region x > x*, but also on the desorption kinetics observed in the region 0 < x < x*. Such information is essential to predict behavior of pesticides under field conditions and will be a subject of future study.

Accurate determination of the sorption of rapidly degradable and/or hydrolyzable compounds is challenging in the proposed method as well. In the present study, no attempt was made to suppress biotic degradation of pesticides during experiments. Nevertheless, the unsaturated flow method yielded a Kd for dichlorvos that was consistent with the position of pesticide displacement front observed in the column. The method assumes that continued equilibrium is maintained between the aqueous solution phase and the sorbed phase in the region x > x*. However, this assumption becomes questionable if the rates of degradation and/or hydrolysis are different between the two phases. Further studies, including those using biocides, will be required to investigate the degree of perturbation of sorption equilibria by degradation and/or hydrolysis of the compounds.

From the above, we conclude that the unsaturated transient flow method is suitable for measuring sorption of hydrophilic organic compounds having a small value of Kd (<2 L kg–1 or Koc of 200 for soil and/or sediment with organic carbon content of 10 g kg–1). The method is versatile and also has a potential to be utilized for investigating sorption and desorption kinetics at a range of soil water contents encountered in the field. In addition, the method may be used for determining sorption of hydrophilic, rapidly degradable and/or hydrolyzable compounds—a class of compounds that are most difficult to handle with conventional batch methods—although further study is necessary to confirm and substantiate this possibility.


   ACKNOWLEDGMENTS
 
World Bank/Government of Pakistan provided financial support to Dr. Riaz Ahmad for his Doctoral studies. Thanks to Dr. Warren Bond (CSIRO, Canberra) and Dr. Jan Vanderborght (Forschungszentrum Juelich, Germany) for their helpful comments on the manuscript. Dr. Bond kindly loaned us the Perroux tube for the experiments. Critical comments from anonymous reviewers and the technical editor of the journal are also acknowledged.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 THEORY
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 REFERENCES
 




This article has been cited by other articles:


Home page
 Soil Sci.Home page
T. E. Ochsner, B. M. Stephens, W. C. Koskinen, and R. S. Kookana
Sorption of a Hydrophilic Pesticide: Effects of Soil Water Content
Soil Sci. Soc. Am. J., October 27, 2006; 70(6): 1991 - 1997.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (3)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Ahmad, R.
Right arrow Articles by Kookana, R. S.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Ahmad, R.
Right arrow Articles by Kookana, R. S.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by Ahmad, R.
Right arrow Articles by Kookana, R. S.
Related Collections
Right arrow Laboratory Column Studies
Right arrow Pesticides
Right arrow Sorption/Exchange
Right arrow Organic Compounds
Right arrow