Environ. Sci. Technol. 2003, 37, 852-858
Identification and Characterization of Sorption Domains in Soil Organic Matter Using Structurally Modified Humic Acids AMRITH S. GUNASEKARA,† MYRNA J. SIMPSON,‡ AND B A O S H A N X I N G * ,† Department of Plant and Soil Sciences, University of Massachusetts, Stockbridge Hall, Amherst, Massachusetts 01003, and Division of Physical Sciences, University of Toronto, Scarborough College, 1265 Military Trail, Toronto, Ontario M1C 1A4, Canada
The sorption of phenanthrene was examined in humic acids (HAs) from different sources: a compost, a peat soil, and a mineral soil. Sub-samples of each HA were subjected to bleaching or hydrolysis to remove predetermined chemical groups from their structures. Bleaching successfully removed a large percentage of rigid, aromatic moieties, whereas hydrolysis removed the mobile, carbohydrate components. Phenanthrene sorption by all HAs was nonlinear (N < 1). However, the phenanthrene isotherms of the bleached HAs were more linear than those of the untreated HAs, whereas the removal of the carbohydrate components by hydrolysis produced more nonlinear isotherms. The introduction of pyrene to the phenanthrene sorption system yielded more linear isotherms for all the HAs, indicative of competitive sorption. Proton spin-spin (1H T2) relaxation determined by nuclear magnetic resonance (NMR) was used to identify separate rigid (condensed) and flexible (expanded) 1H populations and to determine their distribution. These 1H domains were highly sensitive to temperature and correlated well with reported glass transition temperatures for HAs. In combination with the chemical treatments, sorption, and spectroscopic data, we were able to observe some significant relationships among chemical groups, sorption behavior, and structural characteristics.
Introduction Polycyclic aromatic hydrocarbons (PAHs) are common anthropogenic pollutants in soils and sediments. Soil/ sediment organic matter (SOM) can reduce the risk of PAH exposure because of its strong sorptive affinity. However, the sorption mechanism of PAHs in SOM is not clearly understood as the molecular structure of SOM is extremely complicated and undefined. Different models have been employed to explain the sorptive behavior of PAHs in SOM. Solid-phase dissolution (partition) model is often used (1, 2). It is based on linear isotherms, homogeneous binding-site energy distributions, and noncompetitive sorption processes. However, nonlinear * Corresponding author phone: 413-545-5212; fax: 413-545-3958; e-mail:
[email protected]. † University of Massachusetts. ‡ University of Toronto. 852
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 37, NO. 5, 2003
isotherms (3-6), heterogeneous site energy distribution (7), and competitive sorption between sorbates (8-10) have been observed for SOM. Similar sorption behavior is observed in synthetic polymers (11). Consequently, a number of sorption models, adapted from polymer science theory, were proposed. They are the dual mode sorption model (DMSM) (10, 12), dual reactive domain model (13), and extended dual mode model (14). According to the DMSM (10, 12, and refs therein), sorption in expanded domains, analogous to rubbery polymers, of SOM is through solid-phase dissolution (partitioning), whereas sorption in condensed domains, similar to glassy polymers, is by both partitioning and a hole-filling (adsorption) process. The latter process yields nonlinear and competitive sorption. Although PAH sorption to these domains of SOM is supported by nonlinear isotherms and explained using the multimechanistic models, little is understood about the chemical structure and distribution of these domains in SOM. Several studies have examined the sorption contribution of the condensed and expanded domains. For instance, Xing (15) found a positive correlation between HA sample aromaticity and isotherm nonlinearity, which led to the proposal that the condensed domains may be primarily composed of aromatic moieties. Chin et al. (16) had similar findings when they observed a strong positive correlation between pyrene binding coefficients and aromatic content of aquatic HA. However, a few recent reports (17-19) showed that aliphatic components in SOM could contribute significantly to PAH sorption. For example, Chefetz et al. (19) found that plant cuticle and lignin, precursors of SOM, had nonlinear isotherms and high sorption capacity for pyrene. These contradictory findings in the sorption literature are an issue that must be addressed because enhanced or reduced sorption of PAHs in soil is dependent on the structural composition and conformation of SOM. In this study, we attempt to shed some light on this issue. Four aspects make this study unique to the SOM sorption literature. This will be the first report to use structurally modified HAs to probe the chemical group contributions to sorption domains of SOM. Second, not many studies have used competitive sorption experiments to understand PAH sorption in SOM although they can provide valuable information. For example, competitive sorption experiments were used to determine the occurrence of plasticization (swelling) in SOM (8). Further, to our knowledge no previous reports have examined the relative expanded and condensed domain distribution in SOM using proton spin-spin relaxation nuclear magnetic resonance spectroscopy (1H T2 NMR). This NMR technique was effectively used to examine rigid (condensed) and flexible (expanded) components in coals (20, 21). Finally, this study links PAH sorptive behavior in HA from the sorption experiments to structural information from NMR results. Therefore, we will examine the sorption mechanisms of SOM using single-sorbate (phenanthrene), bi-sorbates (phenanthrene and pyrene), and structurally modified HAs. The results will help us determine the chemical groups of the sorption domains and their role in phenanthrene sorption. We will also estimate the relative distribution of the sorption domains in SOM and determine their behavior in response to elevated temperatures.
Experimental Section Sorbents. Three primary sorbents were used in this study. Humic acid was extracted from a compost, a peat soil, and a mineral soil. Compost HA was extracted from mushroom compost that was prepared using horse manure-straw 10.1021/es026151e CCC: $25.00
2003 American Chemical Society Published on Web 02/04/2003
TABLE 1. Organic Carbon (OC) Contents (%) of Treated and Untreated Compost, Peat, and Soil HAs and Their Structural Carbon Distributions (%) as Obtained from the 13C NMR Integration Results HA source
chemical treatment
HA OC
alkyl C
O-alkyl C
aromatic C
phenolic C
carboxyl + carbonyl C
compost
untreated bleached hydrolyzed untreated bleached hydrolyzed untreated bleached hydrolyzed
56 51 62 55 57 59 49 34 60
20 31 23 24 50 22 17 35 16
39 47 26 28 27 21 29 30 14
22 7 33 24 8 31 29 17 46
10 3 13 9 3 12 9 5 12
9 11 5 15 12 15 16 13 12
peat soil
bedding, poultry manure, gypsum, and brewer’s grain. The organic matter of compost is “young” compared to peat and soil SOM and its properties and chemical group distribution have been reported by Chen et al. (22). Peat HA was extracted from the International Humic Substances Society peat soil. Soil HA was isolated from a mineral soil (Mollisol, Edmonton, AB). The HA extraction procedures followed the methods outlined by Swift (23). Chemical Treatments. Two different chemical treatments were used to remove predetermined structural components from HAs. The HA samples were subjected to hydrolysis, which involved mixing them with 6 M HCl and maintaining the system under reflux for 6 h. Details of this procedure were described elsewhere (24). A second chemical treatment, bleaching, involved treating the HAs with 10 g of sodium chlorite, 10 mL of acetic acid, and 100 mL/g HA of deionized water (25). Following the chemical treatments, the solid residue was separated from the mixture by centrifugation at 4000g for 15 m and dialyzed using a Spectra/Por Membrane (Fisher Scientific) with a molecular weight cutoff of 8000 Da. All treated HAs were freeze-dried, ground, and stored for their characterization and sorption work. Sorbates. Phenanthrene was used as the “primary” sorbate and pyrene was used as the “competing” sorbate. These compounds were chosen because they are common pollutants at many contaminated sites. The [Ring-UL-14C] and unlabeled PAHs (>98% purity) were purchased from Sigma-Aldrich chemical company (St. Louis, MO) and used without further purification. Sorption and Competitive Sorption Experiments. The batch equilibration experimental technique at 20 ( 1 °C (12) was used to obtain the isotherms for all HAs. The isotherms were obtained using 125-mL serum bottles with Teflon-lined septa in aluminum seals. Sorption to the bottles was insignificant. The background solution consisted of 0.01 M CaCl2 and 200 µg/mL of HgCl2 as a biocide. The labeled and nonlabeled phenanthrene were added to the background solution and mixed for approximately 1 h before adding them to the sorbent. Mixing was initiated continuously for 5 days in an incubator shaker (New Brunswick Scientific Series 25) to allow the sorbates to reach apparent sorption equilibrium. Previous HA sorption studies found that no significant increase in sorption was observed after 1 day in a 7-day sorption experiment (3). The bottles were centrifuged and 1 mL of supernatant was added to a ScintiVerse cocktail (12 mL) prior to scintillation counting. All isotherms were obtained using duplicate samples and blanks. Sorbed chemical concentrations were determined by mass balance calculations. To reduce experimental error, the quantity of HA added to the background solution was adjusted to maintain sorption of phenanthrene between 30 and 70%. Phenanthrene concentration points were evenly spaced on a log scale ranging from 0.006 to 0.8 µg/mL and HA used ranged from 4 to 10 mg per bottle depending on the sorption capacity of individual HAs.
Competitive isotherms were obtained by introducing 0.16 µg/mL pyrene in the equilibrated phenanthrene sorption systems (i.e., after 5 days equilibration). Because of the low solubility of phenanthrene (1.15 µg/mL) and pyrene (0.16 µg/mL), they were first dissolved at high concentrations in methanol before being added to the background solution, but the total amount of methanol did not exceed 0.1 vol %. The bottles containing phenanthrene and pyrene were mixed for an additional 5 days. After the solutions were mixed, the bottles were centrifuged and a 1-mL aliquot of the supernatant was sampled according to the procedures described above for scintillation counting. Mass balance calculations were conducted to assess the behavior of phenanthrene in the presence of pyrene. All sorption and competitive sorption data were fitted to the logarithmic form of the Freundlich equation:
log S ) log KF + N log C
(1)
where S is the total sorbed concentration (µg/g), C is the final solution phase concentration (µg/mL), and KF (µg/g) (µg/mL)-N and N are constants with N often
