Partitioning Behavior of Insecticides in Soil-Water Systems: II. Desorption Hysteresis Effects. 1985

This investigation was undertaken to determine whether desorption hysteresis effects were dependent upon methods used in obtaining desorption data. Adsorption-desorption isotherms were obtained for two organophosphorus insecticides in aqueous suspensions of a clay, a sandy loam, and an organic soil. Desorption isotherms were obtained using both the consecutive desorption method and the dilution method. In the consecutive method, the same sample (after adsorption step) was put through a series of equilibration-centrifugation steps in which, following equilibration, the adsorbent was thrown down by centrifugation, and part of the supernatant pesticide solution was replaced by distilled water before reequilibration. In the dilution method a series of replicate samples (same adsorbent weight), after the adsorption step, were diluted to different volumes with distilled water, reequilibrated, then centrifuged to separate the phases. With the exception of one insecticide-soil system (where both methods produced minimal hysteresis), hysteresis effects were considerably reduced by using the dilution method. Repeated centrifugation appeared to be associated with the appearance of hysteresis effects observed using the consecutive desorption method. A short discussion is included on the improper use of desorption data to construct "single point desorption isotherms," which has created confusion in the literature. Additional Index Words: irreversibility, non-singularity, isotherm, organic matter. Bowman, B. T., and W. W. Sans. 1985. Partitioning behavior of insecticides in soil-water systems: Il. Desorption hysteresis effects. J. Environ. Qual. 14:270-273. 1 Contribution from the Research Centre, Canada Agriculture, University Sub Post Office, London, Ontario, N6A 5B7. Received 27 Jan. 1984. 2 Research scientist and research assistant, respectively. The existence of nonsingularity, or hysteresis effects, in adsorption-desorption systems involving pesticides has been reported by numerous researchers (2, 5, 10, 17, 19). These effects manifest themselves when a desorption isotherm has a lesser slope than its respective adsorption isotherm, indicating a partial irreversibility. In pesticide-soil-water systems exhibiting hysteresis effects, significant over-estimates of pesticide desorption could be made if predictions were based only on the adsorption isotherm parameters. Rao and Davidson (17) have identified three major possible sources of hysteresis effects as (i) artifacts created by the particular methodology, (ii) failure to establish complete equilibrium during the adsorption period, and (iii) chemical and/or microbial trans-formations of the pesticide during the experiment. Rao et al. (18) found that the centrifugation-resuspension step in the batch-type adsorption-desorption method might be responsible for at least part of the hysteresis, and instead suggested using a dilution method. However, Horzempa and DiToro (10), using a variation of the dilution method, reported that centrifugation did not appear responsible for nonsingularity effects in their studies. Peck et al. (16) reported that as organic matter content of the adsorbent increased, hysteresis became more pronounced for diuron adsorption-desorption processes. This may be, in part, explained by a secondary slow adsorption process by organic matter. Khan (11) reported that 2,4-D and picloram adsorption by humic acid was slow in reaching equilibrium. We have noted (unpublished data) that chlorpyrifos exhibited a further 15% adsorption by an organic soil in a second 24-h adsorption period. The entire issue of adsorption-desorption hysteresis has been unnecessarily complicated and confused by the improper presentation of desorption data in the form of "single-point desorption isotherms" (5, 7, 9, 10, 20). These graphical representations of desorption have been created by joining the first desorption points of separate systems (samples), not recognizing the thermodynamic singularity of the desorption process with respect to its initial starting point. This type of desorption representation can be easily spotted because the "desorption" branch has a steeper slope than the adsorption branch, implying that desorption occurred more readily (in a thermodynamic sense) than the adsorption process. The purposes of this report are to describe our study of the effect of two desorption methodologies observed in several insecticide-soil adsorption-desorption systems, directed toward the elucidation of the causes of this phenomenon, and to discuss the graphical representation of desorption data. MATERIALS AND METHODS The chemical names of substances referred to are given in the Appendix. The purity of parathion was 98.8% and fensulfothion sulfone was recrystallized from methanol at -20oC.Air-dried samples (< 40 mesh) of Bondhead sandy loam (A horizon, 3.9% organic matter), a Ca-saturated Morris illite, and an organic soil (36.7% organic matter), the properties of which have been previously described (3), were used as adsorbents. Adsorption isotherms were constructed using the batch method and 30-mL volumes. Adsorbent weights for both adsorption and desorption isotherms were approximately 0.5, 0.75, and 1.0 g for the Ca-illite, organic soil, and Bondhead systems, respectively. All adsorption and desorption equilibration times were 24 h. For the initial adsorption step in both desorption procedures, triplicate 30-mL volumes of either a 10 or 30 μg mL solution of parathion or fensulfothion sulfone, respectively, were equilibrated with the appropriate amount of adsorbent. In the consecutive desorption method, samples were equilibrated in 150-mL Corexcentrifuge tubes. After centrifugation (6000 x g, 30 min) a volume of supernatant was removed, replaced by an equal volume of distilled water, and reequilibrated. This supernatant replacement procedure was repeated a total of four times, with the removed aliquot being re-centrifuged (43,500 x g, 10 min) to remove any remaining adsorbent traces. In the dilution desorption method, 15 samples were put through the adsorption step, using the 30-mL volume. After equilibration, one triplicate set was centrifuged (43,500 x g, 15 min) and the supernatant concentration was determined. The remaining 12 samples (four sets of triplicates) were diluted with different volumes (60 to 370 mL) of distilled water (final volumes, 90-400 mL), and reequilibrated. Appropriate aliquots of the suspension were centrifuged (43,500 x g, 15 min) and the supernatant concentration determined. All experiments were conducted at 20 ± 1.5oC. With both desorption techniques, desorption points at four equilibrium concentrations were obtained after the initial adsorption ________________________________ 3 See the Appendix for a listing of the chemical names of pesticides used in this article.