Contaminants of environmental concern commonly reside in the sediment or solid phase. The extent and rate of desorption has heretofore been particularly unpredictable. In the present research, the adsorption and desorption of seven organic compounds with water solubilities ranging from 0.005 to 517 mg/L have been studied in natural sediments. In every case, a fraction of the adsorbate was adsorbed irreversibly (i.e., desorption was not the opposite of adsorption, yet the sorbate is not covalently bonded to the sediment). Each sediment-contaminant combination exhibited a fixed maximum irreversible adsorption, , which could be filled in one or several steps and which is related to common molecular properties and sediment organic carbon content (OC). For most compounds, (μg/g) ≈ 103.8OC. Furthermore, the OC-normalized partition constant for this irreversible compartment is es sentially constant for the compounds and sediments studied with = 105.53±0.48 mL/g. After about 1−3 days of contact time, all laboratory adsorption and desorption data could be modeled using a single isotherm equation, based upon commonly measured chemical and sediment parameters. The isotherm equation consists of two terms, a linear term to represent reversible sorption and a Langmuirian-type term to represent irreversible sorption. This combined isotherm is used to interpret numerous published field studies. The potential impact of this model on sediment quality criteria (SQC) and remediation are discussed.

Several unique features of sorption irreversibility have been investigated in this paper. Adsorption has been found to be biphasic with about 30−50% of the adsorbed mass residing in the irreversibly sorbed compartment, until this compartment is filled, and the rest of the mass resides in the labile compartment. Naphthalene in the reversible compartment follows a linear adsorption isotherm with a normal organic carbon-based partition coefficient. A finite fixed total compartment size is observed for the irreversible fraction, (μg/g), on both natural and surrogate solids. In multiple batch adsorption/desorption experiments, the maximum concentrations that resist desorption are ≈ 10 μg/g for naphthalene on Lula sediment and ≈ 0.36 μg/g for 2,2‘,5,5‘-tetrachlorobiphenyl (2,2‘,5,5‘-CB) on both Lula and surrogate solids. The concentration in the irreversibly sorbed compartment varied with the initial naphthalene concentration available for adsorption. In addition, the amount in the irreversibly sorbed compartment increases linearly with the number of adsorption steps until the maximum concentration is reached. After the maximum concentration of the irreversibly sorbed compart ment is satisfied, the adsorption/desorption of naphthalene and 2,2‘,5,5‘-CB becomes reversible. The irreversibly sorbed compartment appears to be at equilibrium with the aqueous phase when the labile naphthalene or 2,2‘,5,5‘-CB is removed, but the equilibrium concentration is much lower than would be predicted with conventional hydrophobic partitioning theory. The aqueous phase concentration in equilibrium with the irreversibly sorbed compartment is about 2−5 μg/L for naphthalene and 0.05−0.8 μg/L for 2,2‘,5,5‘-CB. Similar adsorption/desorption phenomena are observed with both a natural sediment and a well-characterized sorrogate solid.

Competitive sorption to natural solids among mixtures of organic compounds has been documented in the literature. This study was conducted to determine co‐solute competitive effects on the biological and physical availability of polycyclic aromatic hydrocarbons in soils after long contact periods (aging). Sterile suspensions of Mount Pleasant silt loam (Mt. Pleasant, NY, USA) and Pahokee peat soils were spiked with phenanthrene and allowed to age for 3 or 123 d before inoculation with a phenanthrene‐degrading bacterium in the presence or absence of the nonbiodegradable co‐solute pyrene. As expected, mineralization decreased with aging in the samples not amended with pyrene. However, addition of pyrene just prior to inoculation at 123 d significantly mitigated this decrease; that is, the extent of mineralization was greater in the 123‐d pyrene‐amended samples than in the 123‐d nonamended samples. Parallel experiments on sterile soils showed that pyrene increased the physical availability of phenanthrene by competitive displacement of phenanthrene from sorption sites. First, the addition of pyrene increased recovery of 123‐d‐aged phenanthrene by mild solvent extraction. Second, addition of pyrene (at three concentrations) dramatically reduced the apparent distribution coefficient (Kappd) of several concentrations of 60‐, 95‐, and 111‐d‐aged phenanthrene. At the lowest phenanthrene and highest pyrene concentrations, reductions in the Kappd of phenanthrene in the peat soil reached 83%. The competitive displacement effect observed in this study adds further support to the dual mode model of sorption to soil organic matter. The displacement of an aged contaminant by a nonaged co‐solute might also prove useful in the development of novel remediation strategies.

A surrogate sediment was developed to reduce some of the complexity in the structural aspects of the adsorbed organic carbon phase. Layers of an anionic surfactant were sorbed to colloidal anatase to produce an organic carbon phase that had hydrophobic regions and resisted desorption. The surrogate is verified as a model sediment by comparing the results of contaminant [2,2′,5,5′-tetrachlorobiphenyl (PCB) and naphthalene] adsorption and desorption batch experiments to the results of similar experiments performed on a well-studied natural sediment. The surrogate exhibited adsorption of contaminants via a hydrophobic interaction in the same magnitude as to a natural sediment for the two different hydrophobic organic contaminants in 0.1 or 0.15 M NaCl solution. The sorption of PCB to the surrogate at varied organic carbon contents was observed to follow a linear adsorption isotherm. Desorption experiments were conducted by successive dilutions. Both the surrogate and natural sediment were observed to exhibit similar desorption behavior. The solution concentration during desorption was lower than predicted by the adsorption isotherm and remained unchanged from 4 h to 168 days. The heterogeneous nature of sediments should be greatly reduced in the surrogate yet desorption still appears to be low.

The bioavailability of an organic compound in a soil or sediment commonly declines with the soil‐chemical contact time (aging). A series of parallel desorption and bioavailability experiments was carried out on phenanthrene previously aged up to ∼100 d in Mount Pleasant silt loam (Mt. Pleasant, NY, USA) or Pahokee peat soil to determine as a function of the aging period the degree of correlation between the reduction in bioavailability and the rate and extent of desorption and the influence of soil organic matter composition on availability. The mineralization of phenanthrene by two bacteria and the uptake of phenanthrene by earthworms showed expected declines with aging. Likewise, the rate of phenanthrene desorption in the absence of organisms decreased with aging. The decline in initial rate of mineralization or desorption was nearly an order of magnitude after 50 to 60 d of aging. Plots of normalized rates of mineralization or desorption practically coincided. Similarly, plots of normalized fraction mineralized or fraction desorbed during an arbitrary period gave comparable slopes. The partial removal of organic matter from the peat by extraction with dilute NaOH to leave the humin fraction reduced the biodegradation of phenanthrene aged for 38 and 63 d as compared to the nonextracted peat, but the effect disappeared at longer incubation times. The rate of desorption from samples of peat previously extracted with NaOH or Na4P2O7 declined with aging and, for a given aging period, was significantly slower than from nonextracted peat. This work shows that the reduction in bioavailability of phenanthrene over time in soil is directly correlated with reduction of its physical availability due to desorption limitations. In addition, this study shows that removal of extractable humic substances leads to a decline in the rate of desorption and in the bioavailability of the substrate.