Laboratory batch equilibrium studies were conducted to evaluate the sorption-desorption behavior of imazethapyr and its major plant metabolite, 5-hydroxyimazethapyr, in three Minnesota soils. Sorption of both compounds on all soils was low, and pH did not significantly influence sorption in the range of 4,8 to 7.1. Less 5-hydroxyimazethapyr was sorbed than imazethapyr on the same soil. Once sorbed, both compounds were only partially desorbable from all soils. Significant hysteresis and formation of nonextractable residues indicate that the small amount of chemical sorbed is bound to selective surfaces with strong bonds. Webster clay loam had greater irreversibility than Waukegen silt loam and Estherville sandy loam soil. The hysteresis observed in desorption may be responsible for the difference between mobility estimations made from laboratory sorption studies and the limited mobility observed in field experiments.

Adsorption and desorption isotherms for atrazine and two metabolites, hydroxyatrazine (HA) and S-glutathione atrazine (GSHA), were determined by batch equilibration on Plano and Waukegan silt loam soils at two soil pH levels (Plano, 6.1 and 4.5; Waukegan, 6.1 and 4.0). Freundlich adsorption isotherms were not affected by soil type except for GSHA at pH 4.0 to 4.5. When averaged over both soils, the order of adsorption at pH 6.1 was atrazine (Kf = 3.7) < GSHA (Kf = 7.3) << HA (Kf = 25) and at pH 4.0–4.5 was atrazine (Kf = 6.1) << HA (Kf = 58) ≤ GSHA (Kf: Plano = 35; Waukegan = 78). The average slope of the adsorption isotherms (1/nads) was 0.81. The slopes of all desorption isotherms (1/ndes) were less than their respective 1/nads, indicating hysteresis. Atrazine desorbed into soil solution (1/ndes > 0.0). With the exception of GSHA which desorbed from the pH 6.1 Plano silt loam (1/ndes = 0.15), desorption of HA and GSHA from other treatments was negligible (1/ndes = 0.0). Consequently, leaching of HA and GSHA in these and similar soils is not likely, due to high adsorption and low desorption.

Herbicide desorption isotherms may be affected by the amount of nondesorbable herbicide present in soil. Nondesorbable alachlor (as determined after methanol extraction) generally increased on a Waukegan silt loam (Typic Hapludolls) and a Ves clay loam (Udic Haplustolls) during five 0.01 M CaCl2 desorptions. Atrazine was totally extracted with methanol from the Waukegan soil after one desorption using 0.01 M CaCl2. However, after five desorptions with 0.01 M CaCl2 an average of 5.5 and 15.5% of the total recovered atrazine from two atrazine application rates was methanol nondesorbable from the Waukegan and Ves soils, respectively. Freundlich desorption isotherms adjusted for nondesorbable herbicide accounted for as much as 71 % of the difference between adsorption and desorption isotherms. Only a portion of the hysteresis observed can be attributable to nondesorbable herbicide.

The purpose of this study was to quantify hysteresis during adsorption and desorption of atrazine as a function of incubation time for a Sharkey clay soil. Adsorption was carried out using one day batch equilibration and was followed by incubation periods ranging from 1 to 24 d. Incubation was subsequently followed by six consecutive desorption steps where each step represented 1 d. The Freundlich equation (S = K CN where S is the amount of atrazine retained, μg g-1; C is concentration, μg ml-1; K is the distribution coefficient, cm3g-1; and N is a dimensionless parameter) was used to describe batch results. Both adsorption and desorption isotherms were well described by the Freundlich model. Fitted K parameter values for desorption isotherms were consistently higher than those associated with adsorption. The opposite trend was observed for the exponential parameter N. The results revealed that desorption deviated significantly from adsorption data. The deviation, which is commonly referred to as hysteresis, was more pronounced as incubation time increased. Batch equilibration results also indicated that the extent of hysteresis was not influenced by soil sterilization. Attempts to quantify the extent of hysteresis using a simplified approach are presented. We found that, for a given batch data set, hysteresis can be quantified provided that Freundlich N from adsorption and desorption isotherms is known.