JEQ Grow Your Career With ASA
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

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 Related articles in JEQ
Right arrow Similar articles in this journal
Right arrow Similar articles in 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 Google Scholar
Google Scholar
Right arrow Articles by Smith, M. C.
Right arrow Articles by Kingery, W.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Smith, M. C.
Right arrow Articles by Kingery, W.
Agricola
Right arrow Articles by Smith, M. C.
Right arrow Articles by Kingery, W.
Related Collections
Right arrow Agricultural Pesticides
Right arrow Turfgrass Pesticides
Right arrow Laboratory Column Studies
Right arrow Pesticides
Right arrow Nonequilibrium Transport
Right arrow Organic Compounds
Published in J. Environ. Qual. 33:581-593 (2004).
© ASA, CSSA, SSSA
677 S. Segoe Rd., Madison, WI 53711 USA

TECHNICAL REPORT

Organic Compounds in the Environment

Incubation Time Effects on Imazaquin Desorption as Determined by Nonequilibrium Thin-Soil Disc Flow

M. Cade Smith*, David R. Shaw, Joseph H. Massey, Michele Boyette and William Kingery

Department of Plant and Soil Sciences, Box 9555, 117 Dorman Hall, Mississippi State University, Mississippi State, MS 39762

* Corresponding author (csmith{at}pss.msstate.edu).

Received for publication September 24, 2002.

   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Because organic sorption in soil may never reach equilibrium, a thin-disc flow nonequilibrium method may be helpful in understanding herbicide–soil interactions. This research was conducted to (i) determine the influence of incubation time on imazaquin [2-(4,5-dihydro-4-methyl-4-(1-methylethyl)-5-oxo-1H-imidazol-2-yl)-3-quinolinecarboxylic acid] desorption from soil, (ii) examine the influence of solution flow velocities on desorption, and (iii) elucidate the most appropriate kinetic model to describe imazaquin leaching. Soil at 7.5% moisture w/w was treated with imazaquin and incubated for 24, 72, and 168 h. Treated soil was sealed in an in-line filter apparatus and rinsed with 5.0 mM CaCl2 at 0.33, 0.67, or 1.0 mL min–1. Effluent was collected as 1.0-mL fractions for a total of 50 mL. Flow was stopped for 24 h. When flow resumed, fractions were collected for an additional 15 mL. After the initial desorption, 79% of the imazaquin incubated for 24 h was leached. Increasing incubation time beyond 24 h reduced imazaquin leaching. After both desorption events, 13% of the initially applied imazaquin remained in the soil incubated for 168 h, compared with 7% with soil incubated for 24 h. Elovich and Freundlich kinetics accounted for 98% of the variance observed in the imazaquin desorption curves. First-order and diffusion kinetics accounted for 91% of the variance. Incubating soil for 72 h before desorption reduced the rate of imazaquin desorption by approximately 12%, compared with the 24-h incubation treatment. Imazaquin desorption was not affected by wash solution flow rate. These data suggest that the kinetics of desorption in prolonged desorption events are limited by transport phenomena (i.e., particle and film diffusion).


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
THE KINETICS OF pesticide sorption and desorption have received little attention, as most partitioning studies have focused mainly on the equilibrium aspects of pesticide–soil interactions (Rao and Jessup, 1983; Yaron et al., 1985; Kookana et al., 1992). Pesticide–soil interactions are customarily determined using batch equilibrium experiments. The solution–soil mixture is often allowed to equilibrate from 4 to 24 h (Leistra and Smelt, 1981; Nicholls et al., 1982). A single partitioning coefficient is usually calculated from these experiments. As reported by Ball and Roberts (1991a)(1991b), this partitioning coefficient is then extrapolated across a length of time and may result in poor prediction of how herbicides will behave in the field in either a relatively rapid reaction (i.e., leaching through a soil profile) or a long-term reaction (i.e., sorbing to or desorbing from aged herbicide residues from repeated herbicide applications).

Pesticide sorption kinetics may also be determined by traditional batch techniques. The time of herbicide exposure to soil is varied to measure how herbicide sorption changes over time (Boesten and van der Pas, 1988; Ball and Roberts, 1991a, 1991b; Dang et al., 1994; DiVincenzo and Sparks, 1997; Moreau and Mouvet, 1997). The difficulty in using traditional batch techniques to measure short-term sorption kinetics arises in separating the batch solution from the soil suspension. Because of the time of centrifugation, it is difficult to determine rapid sorption and kinetics for time periods of less than 5 min (Wahid and Sethunathan, 1978; Ball and Roberts, 1991b). However, the batch method is well suited for longer experiments.

To avoid the confounding factor of centrifugation time, a variety of nonequilibrium sorption techniques have been examined. These include the miscible-displacement method (Sparks et al., 1980; Brusseau et al., 1991; Gamerdinger et al., 1991), stirred slurry filtration method (Lindstrom et al., 1970; Wahid and Sethunathan, 1978; Li et al., 1996), and thin-soil disc flow method (Skopp and McCallister, 1986; Miller et al., 1989; Thabet and Selim, 1996). Each of these methods allow for the sampling of solution and determination of herbicide sorption without centrifugation. Thus, nonequilibrium methods may better determine rapid-reaction kinetics than traditional batch methods.

It may be most appropriate to think of herbicide sorption as a multiphasic reaction. Research demonstrates that sorption is composed of a rapid, almost instantaneous phase, followed by a more gradual phase (Hamaker et al., 1966; Letey and Oddson, 1972; Chien and Clayton, 1980; Boesten and van der Pas, 1988; Ball and Roberts, 1991b; Kookana et al., 1992; O'Dell et al., 1992; Dang et al., 1994; Chardon and Blaauw, 1997; Johnson et al., 2000). The rapid phase is most likely the result of sorption on surface sites of organic matter, clay colloids, and soil organic matter colloid complexes (Ball and Roberts, 1991b). The rapid phase of sorption is thought to be reversible, resulting in sorption and desorption partitioning coefficients that are practically equal.

The more gradual phase of sorption probably results from diffusion of herbicides into three-dimensional soil structures. Herbicide sorption experiments conducted over several days have demonstrated that sorption coefficients are not constant, but rather increase with time (Lehman et al., 1990; Pignatello and Huang, 1991; Scribner et al., 1992). Additionally, diffusion-driven reactions result in sorbed herbicide molecules that are thought to be less available for desorption than the herbicide molecules simply sorbed to surface sites. The rate at which diffusion-driven reactions reach equilibrium is inversely proportional to the strength of solute sorption (Ball and Roberts, 1991b).

Nonequilibrium sorption can be the result of diffusion-limited transport of pesticide molecules through soil structures (Brusseau and Rao, 1989.) These structures include, but are not limited to, the interstitial micropores of soil aggregates, interstitial colloid spaces, and three-dimensional matrices of natural organic matter (Wu and Gschwend, 1986, 1988; Bouchard et al., 1988; Nkedi-Kizza et al., 1989; Pignatello, 1990; Ball and Roberts, 1991a, 1991b).

Kinetics of herbicide–soil interactions have been described by a variety of empirical and mechanistic models. These include simple one-site linear or nonlinear kinetic models, two-site sorption models, and several others (Nielsen et al., 1986). The array of kinetic equations used to describe kinetics of soil–herbicide interactions includes zero-, first-, and second-order (Sparks, 1989); Elovich (Chien and Clayton, 1980; Chien et al., 1980; Torrent, 1987); Freundlich (Cooke, 1966; Barrow and Shaw, 1975; Evans and Jurinak, 1976; Torrent, 1987; Elkhatib and Hern, 1988); and diffusion (Crank, 1975).

Imazaquin is a weakly acidic herbicide with a pKa of 3.8, and is moderately sorbed in the nondissociated form but weakly sorbed in the anionic form (Mangels, 1991). Imazaquin sorption is negatively correlated with soil pH (Basham et al., 1987; Renner et al., 1988; Loux et al., 1989; Loux and Reese, 1992). At a solution pH of 4.5, a Hoytville clay soil sorbed 35% of the imazaquin initially in the soil solution. However, with a pH of 5.5 and greater, sorption decreased to less than 10% (Loux and Reese, 1992). The reported Kd values for imazaquin range from 0.049 to 2.76 (Renner et al., 1988). Organic matter and clay content also greatly influence imazaquin sorption within a moderate pH range.

The research presented here examines desorption of surface applications of imazaquin to field-moist soil. Batch slurry experiments have shown low imazaquin sorption to soil, especially at near-neutral soil pH, where imazaquin is predominantly negatively charged (Renner et al., 1988). However, these studies are poorly representative of normal field conditions. Under field conditions, the solution to soil ratio will not reach 2:1, unless the soil particles are suspended in surface water. Additionally, herbicides applied under field conditions often interact with field-moist soils for several days before an activating rainfall.

Given that pesticide–soil interactions may never reach a true equilibrium and pesticide sorption–desorption under saturated flow conditions may only have a few seconds to a few minutes to interact with the soil matrix, a thin-disc flow nonequilibrium method may better resolve herbicide–soil interactions than traditional batch-equilibrium methods. The objectives of this research were to (i) determine the influence of incubation time after herbicide application on desorption of imazaquin from field-moist soil in a thin-disc flow system, (ii) examine the influence of solution flow velocities on desorption of imazaquin, and (iii) elucidate the most appropriate kinetic model to describe imazaquin leaching from a surface application of imazaquin.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Soil from the upper 8 cm of a Demopolis silt loam (loamy-skeletal, carbonatic, thermic, shallow Typic Udorthent; 8% clay, 6.2% organic matter, well-drained, pH = 7.6) was used in all experiments. The soil was collected at the Black Belt Branch Experiment Station near Brooksville, MS, from a newly tilled area that had been in continuous hay production for 6 yr before tillage. This area was selected to avoid any previous exposure to imazaquin that may have confounded experimental results. After soil was collected, it was air-dried and sieved through a 1.7-mm sieve.

Imazaquin–soil interactions were examined by using uniformly ring-labeled 14C-herbicide. Imazaquin purity (99.5%) and specific activity (1343 kBq mg–1) were confirmed by high-performance liquid chromatography analysis and liquid scintillation analysis. The relative amount of imazaquin parent material present after incubation and imazaquin degradation products were also determined by HPLC analysis after each experiment.

Twelve grams air-dried soil were placed in three 50-mL polypropylene centrifuge tubes. The soil was wetted with 0.9 mL of 5.0 mM CaCl2 by adding 100-µL increments and shaking the soil mixture after each incremental addition. This brought the soil moisture to 7.5% w/w (dry weight equivalent). All tubes were placed on a shaker with 12-cm horizontal travel and 260 oscillations per minute. Tubes were shaken for approximately 24 h at 21°C. The result was an evenly moistened soil with aggregates no larger than the original 1.7-mm sieved air-dried soil.

After moistening, each soil was treated with 100 µL radiolabeled 2.6 mM imazaquin ammonium-salt solution with a specific activity of 77 kBq mL–1 in 5.0 mM CaCl2. The imazaquin solution was applied directly to the side of the centrifuge tube in 25-µL increments. The tubes were vigorously shaken after each incremental addition to evenly distribute imazaquin throughout the treated soil. The total 14C-imazaquin added to each tube containing soil equaled 0.64 kBq g–1 air-dried soil. The herbicide-treated soil was then placed back on the horizontal shaker for 24-, 72-, and 168-h incubation times.

After incubating the soil for the desired time, 3.2 g of moistened, treated soil from each of the three tubes was placed on separate 0.45-µm hydrophilic mixed cellulose esters filters, secured with the manufacturer-supplied silicon o-rings, and sealed in separate inline filter apparatuses (330 In-Line Filter Holder; Nalge Nunc International, Rochester, NY). A schematic representation is given in Fig. 1 . This created a thin disc of treated soil approximately 3.0 mm deep with a 44-mm diameter. Additionally, six 0.5-g samples of moistened, treated soil from each tube were combusted in a biological oxidizer for 4 min. The evolved 14CO2 was trapped in 15 mL of scintillation cocktail (Carbon 14 Cocktail; R.J. Harvey Instrument, Hillsdale, NJ). Each centrifuge tube was rinsed with three 1.0-mL volumes of methanol to account for any retention of imazaquin by the polypropylene tube. Each methanol rinse was analyzed by liquid scintillation counting.



View larger version (22K):
[in this window]
[in a new window]
  		Fig. 1. Schematic of thin-soil disc flow system.

 
Each disc of soil was rinsed with 5.0 mM CaCl2 at a flow rate of 1.0 mL min–1. Additionally, the 72-h incubation time was rinsed at flow rates of 0.33 and 0.67 mL min–1. All incubation-time and flow-rate treatments were conducted in triplicate. Effluent was collected as 1.0-mL fractions in 20-mL liquid scintillation vials. After collecting 50 mL of effluent, flow was stopped for 24 h; flow was then resumed and 1.0-mL fractions of effluent were collected for an additional 15 mL. Each 1.0-mL fraction was combined with 10 mL water-accepting liquid scintillation cocktail (Ultima Gold; Packard Instrument Co., Downers Grove, IL).

The amount of imazaquin contained in each fraction was determined by liquid scintillation analysis. Additionally, after each soil disc was leached, each component of the experimental apparatus was rinsed with methanol and the filter and soil disc were combusted to account for the mass balance of applied imazaquin. One leachate sample from each incubation time was analyzed by HPLC for radioactive parent imazaquin and degradation products.

The imazaquin–soil concentration during leaching was calculated by subtracting the amount of imazaquin in each fraction from the amount of imazaquin initially in the treated soil. Where appropriate, desorption data were analyzed as split-plot arrangement of treatments. In the experiment examining the influence of incubation time on imazaquin desorption, incubation time was the main-plot factor, and the amount of herbicide in each 1.0-mL fraction was the subplot factor. In the flow rate study, flow rate was the main-plot factor, and amount of herbicide in each fraction was the subplot factor.

First-order, Elovich, Freundlich, and diffusion kinetic models were fit to the observed desorption data within each replicate by calculating the least squares fit through observed data points using the respective equations for each kinetic model. The goodness of fit for each kinetic model was determined by comparing the average R2 values for each model. This was confirmed by plotting the residuals of each model. Additionally, the influence of incubation time and flow rate on the rate of desorption was determined by analysis of variance of the kinetic parameters. In models where slope comparisons are made, linear regressions were fit to the observed desorption data for each replication of each treatment by calculating the least squares fit through observed data points. Differences in regression slopes were determined by analysis of variance of the slope value, and means were separated at the 0.05 level of significance.


   RESULTS AND DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
CONCLUSIONS
 REFERENCES
 
Biological combustion of treated soil before desorption accounted for 82% of the applied imazaquin in the incubation-time treatments (Table 1). In the flow-rate treatments, 82% of the applied imazaquin was present in the 1.0 mL min–1 treatment, and 92% was present in the 0.67 and 0.33 mL min–1 treatments. Thus, the amount of radiolabeled imazaquin concentration was 0.53 kBq g–1 air-dried soil for all treatments with a flow rate of 1.0 mL min–1 and 0.59 kBq g–1 in the treatments with a flow rate less than 1.0 mL min–1.


View this table:
[in this window]
[in a new window]
  		Table 1. Amount of imazaquin fortified and recovered from soil (Demopolis silt loam, 7.5% moisture w/w, pH 7.6) and the centrifuge tube, and soil concentration of applied imazaquin.

 
Liquid scintillation analysis of the centrifuge tube methanol rinses confirmed that the remainder of radiolabel imazaquin adhered to the interior of the centrifuge tubes (Table 1). The total amount of imazaquin contained in each treated soil disc in the incubation study and the 1.0 mL min–1 flow rate treatment in the flow rate study equaled a surface application of 109 g acid equivalent (ae) ha–1 on a treated-surface area basis, evenly diluted through the surface 3.0 mm. This corresponds to an imazaquin–soil concentration of 17.8 µmol kg–1 (Table 1). The herbicide concentration in the 0.67 and 0.33 mL min–1 flow rate treatments corresponded to a surface application of 122 g ae ha–1 imazaquin and equaled a 20.0 µmol kg–1 soil concentration. The field use rate of imazaquin is 133 g ae ha–1. Parent imazaquin accounted for approximately 97% of the observed radioactivity in leachate after all incubation times, and did not differ across treatments (Table 2). Thus, imazaquin degradation during the experimental time was negligible.


View this table:
[in this window]
[in a new window]
  		Table 2. Recovery of applied imazaquin from different experimental components and percentage of radioactivity in leachate as imazaquin parent.{dagger}

 
Because of the difference in imazaquin–soil concentration between the treatments with 1.0 mL min–1 flow rate and those with lower flow rates, all data were converted to percent of applied imazaquin. Henceforth, the percent of applied imazaquin refers to the amount of imazaquin confirmed by biological oxidation to be present in the soil before desorption.

Figure 2 shows the desorption-curves for the 24-, 72-, and 168-h incubation times. All three incubation periods follow the same general curvilinear pattern. Curvilinear sorption and desorption reactions are commonly observed in soil sorption reactions (Hamaker et al., 1966; Letey and Oddson, 1972; Chien and Clayton, 1980; Boesten and van der Pas, 1988; Kookana et al., 1992; Dang et al., 1994; Chardon and Blaauw, 1997).



View larger version (20K):
[in this window]
[in a new window]
  		Fig. 2. Influence of incubation time and fraction of effluent on desorption of imazaquin from soil. Initial imazaquin–soil concentration was 17.8 µmol kg–1. Flow rate for all treatments was 1.0 mL min–1.

 
Approximately 50% of the applied imazaquin desorbed during the first 10 mL of flow with all three incubation times (Fig. 2). The weak retention observed on the Demopolis silt loam agrees with results reported by Smith et al. (2003). Imazaquin is moderately sorbed in the nondissociated form but weakly sorbed in the anionic form (Mangels, 1991). Imazaquin sorption is negatively correlated with soil pH (Goetz et al., 1986; Basham et al., 1987; Renner et al., 1988; Loux et al., 1989; Loux and Reese, 1992). The soil pH in our experiments was 7.6. Thus, imazaquin would occur solely in the anionic form and be weakly sorbed, and would readily desorb given our experimental parameters.

The rate of imazaquin desorption gradually decreased as flow was extended past 10 mL (Fig. 2). Allen et al. (1996) reported that phosphorus release rates also decreased with increased elution volumes and time. By 25 mL of effluent, 70% of the imazaquin was desorbed from the 24-h incubation treatment, compared with approximately 60% desorption from the 72- and 168-h incubation times (Fig. 2). After the initial desorption event, 79% of the applied imazaquin had leached from the soil incubated for 24 h (Table 2). When imazaquin-treated soil was incubated for 72 h, the amount of applied imazaquin to leach from the treated soil during the initial desorption event decreased to 73% and did not change when incubation time was extended to 168 h. The peak imazaquin solution concentration in the initial desorption event from soil incubated 24 h was 5.9 µM, compared with approximately 4.2 µM from soil incubated either 72 or 168 h (Table 3). Extending incubation time beyond 24 h clearly reduced the amount of imazaquin available for rapid desorption. These data support the results of Lehman et al. (1990), Pignatello and Huang (1991), and Scribner et al. (1992), who concluded that sorption by soil does not reach equilibrium within 24 h.


View this table:
[in this window]
[in a new window]
  		Table 3. Influence of incubation time and flow rate on peak solution concentration of imazaquin in the initial and second desorption events, and on partitioning coefficient at the beginning of the second desorption event.

 
Herbicide sorption experiments conducted over several days by other researchers have demonstrated that sorption coefficients are not constants, but rather increase with time (Lehman et al., 1990; Ball and Roberts, 1991a, 1991b; Pignatello and Huang, 1991; Scribner et al., 1992). It can be inferred that sorption of imazaquin continued after 24 h of incubation and resulted in less imazaquin desorbing from the 72- and 168-h incubation treatments than the 24-h incubation treatment. O'Dell et al. (1992) and Johnson et al. (2000) showed similar time dependence to imazethapyr sorption.

Brusseau and Rao (1989), Ball and Roberts (1991a)(1991b), Brusseau et al. (1991), O'Dell et al. (1992), and Pignatello and Xing (1996) concluded that diffusion into three-dimensional soil components (i.e., soil aggregates, organic matter, and their complexes) should be the most important processes accounting for slow sorption. Diffusion of herbicide into three-dimensional structures probably results in residues that are less available for desorption. In our experiments, imazaquin in soil incubated for 72 or 168 h had more time to diffuse into more resistant sorption sites and was more resistant to leaching, compared with the soil incubated 24 h. It is also probable that imazaquin sorption was not complete after the 72-h incubation, but that any possible differences between the 72- and 168-h incubation times were masked by experimental variance (Ball and Roberts, 1991b).

When solution flow was stopped after 50 mL of solution had passed through the soil disc, and then resumed 24 h later, imazaquin desorption followed the same curvilinear pattern observed during the initial desorption phase (Fig. 2). The second desorption phase removed 13% of the applied imazaquin for the 24-h incubation treatment (Table 2). For soil incubated for 72 and 168 h, the second desorption removed 15 and 16% of the applied dose, respectively. The second desorption curve featured a very rapid initial change in imazaquin–soil concentration, and as in the initial desorption event, the rate of imazaquin leaching became more gradual as effluent volume increased. This indicates that the rate of imazaquin removal from mass flow of solution through the soil during the initial desorption event exceeded the rate of imazaquin desorption from the soil (Hinz and Selim, 1999). Thus, the system was not at equilibrium after the initial desorption event. As a result, imazaquin desorption continued during the 24-h static period.

Soil combustion after the two desorption events showed that 7% of the applied imazaquin was retained by the soil disc when soil was incubated for 24 h before desorption, compared with 10 and 13% imazaquin retention from the 72- and 168-h incubation times, respectively (Table 2).

Figure 3 shows the nonequilibrium partitioning coefficients, Kd (L kg–1), for each fraction collected during both desorption events after imazaquin was incubated for 24, 72, and 168 h. The nonequilibrium Kd values were calculated by dividing the soil concentration after each fraction by the solution concentration for the respective fraction. The nonequilibrium Kd values increased linearly with increasing effluent for all incubation times during each desorption event. However, incubation time greatly influenced the rate of nonequilibrium Kd change over the desorption events. During the first desorption event, the slope of the nonequilibrium Kd regression of soil incubated for 72 and 168 h did not differ, and increased approximately 0.88 units for every mL of effluent. However, the slope of the nonequilibrium Kd regression of soil incubated for 24 h increased only 0.49 units for every mL of effluent. Therefore, any difference in slope of the Kd regression across either incubation times or flow velocities would be the result of differences related to the change in imazaquin–soil concentration relative to the change in solution concentration.



View larger version (28K):
[in this window]
[in a new window]
  		Fig. 3. Influence of incubation time and effluent on ratio of solution concentration to soil concentration (nonequilibrium Kd). Initial imazaquin–soil concentration was 17.8 µmol kg–1. Flow rate for all treatments was 1.0 mL min–1. The terms a and b denote initial event slopes that are different at the 0.05 level of significance; y and z denote second event slopes that are different at the 0.05 level of significance.

 
The increase in nonequilibrium Kd with increasing effluent volumes is the result of the imazaquin–soil concentration decreasing at a slower rate than the imazaquin solution concentration (Fig. 3). No literature was discovered relating a nonequilibrium partitioning coefficient to the rate of herbicide nonequilibrium sorption or desorption during saturated flow. The disparity in the rate of removal of imazaquin from the soil disc relative to the soil solution was greater in the 72- and 168-h incubation treatment than in the 24-h incubation treatment. This is the result of the rate of solution concentration decreasing faster with each fraction collected than the desorption rate of imazaquin from the soil disc. This is clear evidence that (i) bulk flow of solution through the soil disc exceeded the rate of imazaquin desorption and (ii) 72- and 168-h incubation times resulted in sorbed imazaquin that was more resistant to desorption than the 24-h incubation time.

The rate of change of the nonequilibrium Kd with each fraction of effluent in the second desorption event is also represented graphically in Fig. 3. The slope of the second desorption nonequilibrium Kd for the 72- and 168-h incubation times equaled approximately 2.15, compared with 0.87 with the soil incubated for 24 h. This suggests that during the second desorption event, imazaquin leached from soil incubated for either 72 or 168 h was nearly 2.5 times more resistant to desorption than from soil incubated for 24 h.

The slopes of all nonequilibrium Kd regressions in the second desorption event were nearly two times greater than those in the first desorption event. The 24-h static period after the initial desorption event allowed for the gradual diffusion of sorbed imazaquin residues into the soil solution (Hinz and Selim, 1999). This resulted in a gradual accumulation of imazaquin in the soil solution during the static period (Table 3). The peak effluent concentration after flow resumed to begin the second desorption event was 1.2 µM from the soil incubated for 24 h, compared with 1.5 and 1.6 µM from the soil incubated for 72 and 168 h, respectively. The solution concentration was then rapidly flushed from soil solution when flow was continued throughout the second desorption event. Hinz and Selim (1999) reported similar results with Zn sorption–desorption reactions in a thin-disc flow system.

As would be expected, flushing of accumulated solution imazaquin was much more rapid in the second desorption event than in the initial desorption event. The initial desorption event was dependent on, and limited by, the rate of desorption of imazaquin from soil. However, when the soil was fully wetted after the first desorption, imazaquin accumulated in the soil solution before desorption, and leaching was not initially dependent on rapid desorption.

The total amount of imazaquin recovered from each desorption event and the soil disc accounted for approximately 99% of the applied herbicide (Table 2). The balance of imazaquin was accounted by the solution in the void volume of the experiment apparatus and the filter contained in the inline filter apparatus. Each of these components accounted for no more than 0.5% of the applied imazaquin (Table 2). The total amount of imazaquin recovered from each incubation treatment ranged from 99 to 101% of the applied imazaquin, and did not differ between incubation times (Table 2).

Figure 4 illustrates the desorption curve for the flow rate study. The same general pattern of imazaquin desorption occurred across all flow rates, as was observed in the incubation study. Imazaquin rapidly desorbed during the first 10 mL of the initial event, which accounted for approximately 50% of the applied imazaquin. Desorption rate gradually decreased after 10 mL of effluent was collected, and approximately 70% of the applied imazaquin had desorbed from all treatments during the first 25 mL of effluent. However, flow rate did not influence the total amount of imazaquin desorption. Because desorption curves are equal across all three flow velocities when plotted as a function of effluent volume, these data suggest that rapid reactions were dominant (i.e., equilibrium driven), as seen by Hinz and Selim (1999). When gradual kinetic reactions are the dominant reaction processes, desorption curves from different flow velocities should exhibit different slopes (Hinz and Selim, 1999).



View larger version (20K):
[in this window]
[in a new window]
  		Fig. 4. Influence of flow rate and fraction of effluent on desorption of imazaquin from soil. Initial imazaquin–soil concentration was 17.8 µmol kg–1. Incubation time for all treatments was 72 h.

 
When the peak imazaquin solution concentrations for different flow rates were compared, it appeared that a greater amount of imazaquin initially desorbed from the 0.67 and 0.33 mL min–1 flow rates than the 1.0 mL min–1 flow rate (Table 3). The peak solution concentration in the initial desorption event was 4.3 µM from soil exposed to 1.0 mL min–1 flow. When the flow velocity decreased to 0.67 or 0.33 mL min–1, the peak solution concentration increased to 6.7 and 8.3 µM, respectively. However, this apparent conclusion is difficult to support given that the initial soil concentration for the 1.0 mL min–1 flow rate was 17.8 µmol kg–1, but 20 µM with the 0.67 and 0.33 mL min–1 flow velocities (Table 1). If the data are mathematically adjusted for solution concentration relative to initial soil concentration, the peak solution concentration from soil exposed to a 0.33 mL min–1 flow rate is still greater than that of the 1.0 mL min–1 flow rate. Allen et al. (1996) also reported that greater initial concentrations initially desorbed faster than lower soil concentrations. The findings of Allen et al. (1996) support the evidence presented here that imazaquin desorption is initially very rapid, but becomes kinetically limited when exposed to longer periods of flow.

The second desorption event accounted for approximately 14% of the applied imazaquin and was not affected by flow rate (Table 2). However, the amount of imazaquin retained by the soil disc after both desorption events was influenced by flow rate. The 65 mL of effluent passed through the soil disc represents approximately 30 void volumes. For the 1.0 mL min–1 flow rate, 10% of the applied imazaquin remained in the soil after both desorption events, compared with 6% retention for the 0.67 and 0.33 mL min–1 flow rates. The difference in imazaquin retention after both desorption events with the different flow velocities may have been the result of a longer desorption time for the lower flow rates (Hinz and Selim, 1999). The system was exposed to 65 min of solution flow with the 1.0 mL min–1 flow rate, compared with 97 and 197 min of solution flow with the 0.67 and 0.33 mL min–1 flow velocities. Furthermore, the initial imazaquin–soil concentration with the 1.0 mL min–1 flow rate was 17.8 µmol kg–1, compared with 20 µmol kg–1 with the 0.33 and 0.67 mL min–1 flow rates (Table 1). The differences in time of exposure to solution flow or initial soil concentrations may have resulted in less imazaquin retention with the slower flow velocities. Allen et al. (1996) reported that phosphorus desorption was dependent on both initial soil concentration and time.

The influence of flow velocity on the regression of nonequilibrium Kd values as a function of effluent volume for the initial and second desorption events is shown in Fig. 5 . Unlike incubation time, flow velocity did not affect Kd regressions for the initial or second desorption event. The results reported by Hinz and Selim (1999) can be used to deduce the significance of a change in Kd values with increasing effluent volume. Since the nonequilibrium Kd values in our flow velocity experiment were equal across all flow velocities, the data suggest that rapid, equilibrium-driven reactions were dominant. If gradual reactions were the dominant retention processes, we would expect regression equations for nonequilibrium Kd from different flow velocities to exhibit different slopes (Hinz and Selim, 1999).



View larger version (28K):
[in this window]
[in a new window]
  		Fig. 5. Influence of flow rate and effluent on ratio of solution concentration to soil concentration (nonequilibrium Kd). Initial imazaquin–soil concentration was 17.8 µmol kg–1 in the 1.0 mL min–1 flow rate and 20.0 in the 0.67 and 0.33 mL min–1 flow rates. All treatments were incubated 72 h before desorption. Regression slopes were not different at the 0.05 level of significance.

 
To elucidate desorption kinetics of imazaquin, desorption data from the initial desorption events in the 24-, 72-, and 168-h incubation times were fit to first-order, Elovich, Freundlich, and diffusion kinetics models. Incubation time did not affect the experimental variance of each model, so desorption data were pooled across incubation times.

Figure 6 is the graphical representation of each kinetic model's fit to the initial desorption event. The vertical error bars equal ±1 standard deviation of the mean for each effluent fraction. The Elovich and Freundlich models accounted for at least 98% of the observed experimental variance, compared with 90% with the first-order model and 92% with the diffusion-based model. Chien and Clayton (1980) reported that phosphate sorption failed to conform to a single first-order kinetic equation. Additionally, Dang et al. (1994) found that first-order kinetics only explained 38% of the observed Zn desorption from soil. Several researchers report that Elovich and Freundlich kinetics more accurately model soil sorption and desorption reactions than first-order kinetics (Chien and Clayton, 1980; Elkhatib and Hern, 1988; Havlin et al., 1985; Aharoni et al., 1991; Dang et al., 1994; Allen et al., 1996).



View larger version (30K):
[in this window]
[in a new window]
  		Fig. 6. Influence of kinetics type on accuracy of modeling imazaquin desorbing from soil, averaged over 24-, 72-, 168-h incubation times before desorption. Initial imazaquin–soil concentration was 17.8 µmol kg–1. Flow rate for all treatments was 1.0 mL min–1. Error bars equal ±1 standard deviation within each fraction of effluent.

 
Figure 7 is the graphical representation of the predicted amount of imazaquin leaching with each of the kinetic models. The first-order kinetics model overpredicted the amount of imazaquin leached from soil during the first five fractions of effluent, and underpredicted imazaquin leaching from the 12th to 25th fraction (Fig. 7a). Aharoni et al. (1991) reported that first-order kinetics did not accurately predict phosphate sorption at short or intermediate reaction periods. In our experiments, first-order kinetics predictions did not differ from the observed amount of imazaquin leached from the soil from the 25th to 49th fraction. Aharoni et al. (1991) also reported that first-order models may be more appropriate predictors of sorption–desorption reactions during later stages of soil reactions than the initial or intermediate phases.



View larger version (31K):
[in this window]
[in a new window]
  		Fig. 7. Influence of empirical model and fraction of effluent on predicted leaching of imazaquin from soil. Initial imazaquin–soil concentration was 17.8 µmol kg–1. Error bars equal ±1 standard deviation within each fraction of effluent.

 
The diffusion-based kinetics model had the same general pattern of accuracy as the first-order model. The diffusion model overpredicted imazaquin leaching in the first four effluent fractions and underpredicted imazaquin leaching during the intermediate fractions (Fig. 7d). Additionally, the diffusion-based model overpredicted imazaquin diffusion during the later stages of diffusion.

The Elovich model accurately predicted imazaquin leaching throughout the entire desorption event (Fig. 7b). Like the Elovich model, the Freundlich model more accurately predicted imazaquin leaching than did the first-order or diffusion models (Fig. 7c). However, the Freundlich model underpredicted imazaquin leaching with the first collected fraction. Being that this fraction was observed in our experiments to contain the greatest concentration of herbicide initially desorbed from the soil (Fig. 2), regardless of incubation time or flow rate, underpredicting herbicide concentration in the initial fraction is a liability when determining the accuracy of each model in predicting the amount of imazaquin leached.

Given these observations, the Elovich model was deemed most appropriate for describing imazaquin desorption. Several researchers report that Elovich kinetics accurately model the gradual phase of sorption–desorption reactions by soil (Chien and Clayton, 1980; Elkhatib and Hern, 1988; Aharoni et al., 1991; Dang et al., 1994). Thus, the Elovich model was selected to compare the influence of incubation time and solution flow rate on the kinetics of imazaquin desorption from soil.

Figure 8 shows the graphical representations of the influence of incubation time on imazaquin desorption from soil during the initial and second desorption events as described by Elovich kinetics. During the initial desorption event, imazaquin from the 24-h incubation desorbed at a faster rate than did imazaquin incubated for 72 or 168 h (Fig. 8a). The slope of imazaquin desorption from the 24-h incubation time was –22, compared with –20 and –19 for 72- and 168-h incubation times, respectively.



View larger version (24K):
[in this window]
[in a new window]
  		Fig. 8. Influence of incubation time on (a) initial phase and (b) second phase of imazaquin desorption from soil as described by Elovich model. Initial imazaquin–soil concentration was 17.8 µmol kg–1. Flow rate for all treatments was 1.0 mL min–1. Initial phase volume was 50 mL, after which flow was stopped for 24 h. Second desorption phase included the subsequent 15 mL of effluent. Regression slopes are different at a 0.05 level of significance when indicated by a or b.

 
Conversely, during the second desorption event after the 24-h static period, the rate of imazaquin leaching was higher from the 72- and 168-h incubation times than the 24-h incubation times (Fig. 8b). The slope of imazaquin leaching was –4.9 from the 24-h incubation times, compared with –6.0 and –6.3 for the 72- and 168-h incubation times, respectively. This is a result of a greater amount of imazaquin being retained in the soil after the initial desorption event in the 72- and 168-h incubation times (Table 2). Thus, during the 24-h static period, more imazaquin desorbed from the soil during the longer incubation times than with the 24-h incubation time. These findings are supported by thin-disc research reported by Hinz and Selim (1999). As a result, when the second desorption phase was initiated, a higher solution concentration was observed to leach from the longer incubation times than from the 24-h incubation time.

Figure 9 graphically illustrates the influence of flow rate on imazaquin desorption from soil during the first desorption event. These data are presented in two separate manners. Figure 9a presents the desorption data as a function of time and Fig. 9b presents the data as a function of effluent volume. When the flow rate data are analyzed as a function of time, the rate of imazaquin desorption during the first desorption event was higher with the 1.0 and 0.67 mL min–1 flow rates than the 0.33 mL min–1 flow rate. The slopes of desorption kinetics as a function of time with the 1.0 and 0.67 mL min–1 flow rates were –20 and –19, respectively, compared with –17 with the 0.33 mL min–1 flow rate. Hinz and Selim (1999) also found that when sorption reactions in thin-disc flow experiments are plotted as a function of time, the rate of soil reactions slows with lower flow velocities.



View larger version (27K):
[in this window]
[in a new window]
  		Fig. 9. Influence of (a) time and (b) effluent volume on leaching of imazaquin from soil as described by Elovich kinetics. Initial imazaquin–soil concentration was 17.8 µmol kg–1 with the 1.0 mL min–1 flow rate and 20.0 µmol kg–1 in the 0.67 and 0.33 mL min–1 flow rates. All treatments were incubated for 72 h before desorption. Regression slopes are different at a 0.05 level of significance when indicated by a or b.

 
When the data were analyzed as a function of effluent volume, the imazaquin desorption kinetics were not affected by flow rate. These data suggest that rapid kinetic reactions processes are dominant because the desorption curves are equal across all three flow velocities when plotted as a function of effluent volume. When gradual kinetic reactions are the dominant retention processes, desorption curves from different velocities should exhibit different slopes when plotted as a function of flow velocity.

The same scenario was observed during the second desorption event (Fig. 10) . When analyzed as a function of time, the rate of imazaquin leaching was greatest with the 1.0 mL min–1 flow rate (Fig. 10a). The slope of imazaquin leaching as a function of time was –6.0 with the 1.0 mL min–1 flow rate, compared with –3.8 with the 0.33 mL min–1 flow rate. However, when imazaquin leaching during the second desorption event was analyzed as a function of effluent volume, the rate of imazaquin leaching was equal across all flow rates (Fig. 10b). As discussed earlier, these data suggest that desorption of imazaquin from soil is initially an equilibrium-driven process (i.e., rapid reaction), but diffusion of sorbed imazaquin from soil is probably the rate-limiting process in imazaquin desorption (Brusseau and Rao, 1989; Brusseau et al., 1991; Pignatello and Xing, 1996).



View larger version (26K):
[in this window]
[in a new window]
  		Fig. 10. Influence of (a) time and (b) effluent volume on second phase leaching of imazaquin from soil as described by Elovich kinetics. Initial imazaquin–soil concentration was 17.8 µmol kg–1. All treatments were incubated for 72 h before desorption. Regression slopes are different at a 0.05 level of significance when indicated by a or b.

 

   CONCLUSIONS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
REFERENCES
 
This research demonstrates that gradual-phase imazaquin sorption continues beyond 24 h of incubation. Furthermore, gradual-phase sorption increases imazaquin's resistance to desorption. During desorption, rapid kinetics initially drive imazaquin desorption. However, when solution flow is continued, rapid kinetics cease to drive imazaquin desorption, and are replaced by gradual kinetic processes (i.e., diffusion). Thus, extended sorption reactions are equilibrium-driven and are probably rate-limited by diffusion of imazaquin out of three-dimensional soil structures.

Imazaquin desorption was more accurately described by Elovich and Freundlich kinetics than first-order or diffusion-based kinetic models. Because Freundlich kinetics underpredicted the amount of imazaquin to initially desorb from treated soil, the most accurate model for describing the leaching of imazaquin from moist soil was the Elovich model. The Elovich model demonstrated that the kinetics of imazaquin desorption were greatly influenced by incubation time before desorption. Soil incubated for either 72 or 168 h desorbed more slowly than soil incubated for 24 h.


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


Related articles in JEQ:

This Issue in Journal of Environmental Quality

JEQ 2004 33: 413-418. [Full Text]  




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 Related articles in JEQ
Right arrow Similar articles in this journal
Right arrow Similar articles in 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 Google Scholar
Google Scholar
Right arrow Articles by Smith, M. C.
Right arrow Articles by Kingery, W.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Smith, M. C.
Right arrow Articles by Kingery, W.
Agricola
Right arrow Articles by Smith, M. C.
Right arrow Articles by Kingery, W.
Related Collections
Right arrow Agricultural Pesticides
Right arrow Turfgrass Pesticides
Right arrow Laboratory Column Studies
Right arrow Pesticides
Right arrow Nonequilibrium Transport
Right arrow Organic Compounds

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
The SCI Journals Agronomy Journal Crop Science
Journal of Natural Resources
and Life Sciences Education		 Vadose Zone Journal
Soil Science Society of America Journal Journal of Plant Registrations The Plant Genome