Chemical Extractions Affect the Structure and Phenanthrene

Florida peat sample is from Palm Beach County, FL. Amherst peat soil developed on glaciolacustrine deposits with no specific clay mineral dominating (...
1 downloads 0 Views 163KB Size
Download PDF
Environ. Sci. Technol. 2005, 39, 8333-8340

Chemical Extractions Affect the Structure and Phenanthrene Sorption of Soil Humin KAIJUN WANG AND BAOSHAN XING* Department of Plant, Soil and Insect Sciences, University of Massachusetts, Amherst, Massachusetts 01003

Humin is a major fraction of soil organic matter and strongly affects the sorption behavior and fate of organic contaminants in soils and sediments. This study evaluated four different extraction methods for soil humins in terms of their organic carbon structural changes and the consequent effects on phenanthrene sorption. Solid-state 13C NMR demonstrated that 0.1 M NaOH exhaustively extracted humin and humin extracted with 6 M HF/HCl at 60 °C had a relatively high amount of aliphatic components as compared with 1 M HF-extracted humin. The treatment of 6 M HF/HCl at 60 °C greatly reduced carbohydrate components (50-108 ppm) from humin samples, i.e., more than 50% reduction. In addition, the humin from this 6 M HF/HCl treatment contained relatively more amorphous poly(methylene) domains than the humins extracted by other methods. With the respect to phenanthrene sorption, the linearity of sorption isotherm (N) and sorption affinity (KOC) varied markedly among the humin samples extracted by different methods. The NaOH exhaustively extracted humin had the most nonlinear sorption isotherm and the HF-extracted humin had the lowest KOC. It is concluded that humin samples from different extraction procedures exhibited substantial differences in their organic carbon structure and sorption characteristics, even though they were from the same soil. Therefore, one needs to be cautious when comparing the structural and sorption features of soil humins, especially when they are extracted differently. The 6 M HCl/HF extraction at elevated temperature is not encouraged, due to the modifications of chemical structure and physical conformation of organic matter.

Introduction Soil organic matter (SOM) is a highly heterogeneous material. Its fractions differ greatly in their structure and sorption affinity to organic contaminants (1, 2). To better understand the underlying mechanisms of SOM-organic contaminant sorption interactions, separation and investigation of each fraction of SOM in terms of chemical compositions and sorption properties are required. Soil and sediment humin is usually defined as a fraction of humic substances that is insoluble in aqueous solution under any pH condition. It has a significant aliphatic nature and appears to be the oldest among the humic fractions: fulvic acid (FA), humic acid (HA), and humin (3). Because humin is the major component of SOM, typically representing more than 50% of organic carbon in soils (4, 5), it is believed to dominate the sorption of * Corresponding author phone: 1-413-545-5212; fax: 1-413-5453958; e-mail: [email protected]. 10.1021/es050737u CCC: $30.25 Published on Web 10/06/2005

 2005 American Chemical Society

hydrophobic organic contaminants in soils and sediments. A previous study (6) has reported that more than 50%, typically 70-80%, of bound polychlorinated biphenyls (PCBs) and polycyclic aromatic hydrocarbons (PAHs) are associated with soil humin. Therefore, understanding the sorption of organic contaminants by humin may provide more insights into the mechanisms of organic chemical contamination and soil remediation. However, humin has not gained as much attention as humic acid or fulvic acid. Research on compositions, structures, and sorption properties of humin has been hampered by its limited solubility and high ash content. The application of solid-state nuclear magnetic resonance spectroscopy (NMR) to humic substances has revealed that humin is enriched with aliphatic structure (7). A NMR relaxation study (8) has proposed that aliphatic and aromatic/ carbohydrate carbons exist as spatially distinct domains in a HF/HCl deashed humin sample: aliphatic domains are relatively rigid, whereas aromatic and carbohydrate domains are somewhat mobile. But recent sorption (9) and spectroscopic studies (10, 11) have shown that the condensed domains appear mainly attributed to aromatic components. Among aliphatic carbon components, both mobile and rigid components have also been identified in NMR investigations (12, 13). Moreover, Cuypers and co-workers (14) investigated the composition of mobile and rigid domains of one soil sample and four sediment samples using thermogravietric analysis, pyrolysis-GC-MS, and CP-MAS 13C NMR; they reported that rigid domains in relatively undecomposed SOM were enriched in aliphatic carbons, whereas rigid domains in relatively weathered SOM were enriched in aromatic carbons. The heterogeneous domains in SOM have been used to explain nonlinear sorption isotherms (15, 16). Sorption of hydrophobic organic compounds (HOCs), e.g., phenanthrene, is greatly affected by the chemical structure and composition of humic substances (7, 17). It has also been documented that the capacity and affinity of SOM to retain HOCs vary significantly among samples of different origins and fractions (1, 11, 18). Humin is reported to have higher sorption affinity and more nonlinear sorption isotherms than HA and its source soil (1, 19). On the basis of the notion that expanded organic matter (HA and FA) would be easily extracted as compared with insoluble condensed organic matter (humin) (20), Gunasekara and Xing (21) proposed a novel structural conformation of humin to explain its high sorption affinity and nonlinearity. They and others (22) suggested that SOM macromolecules, especially the first several molecular layers of the amorphous region in close contact with soil mineral surfaces, might rearrange to take a more condensed conformation, which consequently altered its sorption affinity and isotherm nonlinearity. In this respect, humin samples with different content of minerals may differ in the distribution of mobile/condensed domains of SOM; in other words, extraction and enrichment/purification processes are likely to impact the structural conformation of soil humins and their sorption properties. Humin samples used in the literature were often extracted by different methods. The residues after extraction of HA and FA are usually defined as crude humin. The crude humin, in most cases, contains only a small amount of organic matter dispersed in a large volume of inorganic matrixes. In practice, crude humin is often collected after soil is extracted in 0.1 M Na4P2O7 three times (23); in 0.1 M NaOH once, twice, or three times (24-27); or in 0.5 M NaOH once for 24 h followed by washing five times with 0.5 M NaOH (6). It should be noted that not all HA and FA fractions are extracted out due to the limited number of extractions (5). Some studies used VOL. 39, NO. 21, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

8333

the humins derived from exhaustive extraction to determine the variation of HOCs sorption among SOM fractions (1, 28). The main point of the exhaustive method is to repeat 0.1 M NaOH extraction many times to further remove HA and FA from soil; hence, the organic matter left in the soil residue is not extractable by 0.1 M NaOH, which matches the definition of humin. To get organic C enriched humin samples, the crude humin is usually treated with HF to reduce mineral content. One approach is to treat the crude humin with 1.0 M or 2% (v/v) HF solution (23, 29). Another is to digest the crude humin with hot concentrated HF/HCl (30); this procedure is more chemically aggressive than 1.0 M HF, leading to a variety of possible modifications (3). It is anticipated that different extraction methods may result in structural changes; as a consequence, samples obtained from various extractions may have different sorption behaviors. One recent study (31) has evaluated a series of chemical and thermal treatment methods on four carbonaceous materials (CM): a humic acid, a char, a soot, and a heat-treated soot. Those treatments include HCl, HCl/HF, trifluoroacetic acid (TFA), NaOH, acid dichromate, and heat treatment. They concluded that treatment with HCl, HCl/HF, TFA, and NaOH could be applied to soils and sediments to obtain CM enrichment fractions for sorption evaluation, but acid dichromate and heat treatment might not be appropriate due to the altered composition and sorption properties. However, the samples used in their study were highly organic and contained very little amounts of minerals; thus, treatments aimed at removing minerals failed to build any significant influence on the sorption properties of those CM materials. Contrarily, humin usually contains a large amount of minerals, and deashing treatments are often required at the end of extraction; in this case, various approaches for removing minerals may affect differently the structure and sorption characteristics of humin. However, the impact of extraction methods on the organic carbon structure and sorption behavior of humin has not been reported, yet. Accordingly, the objectives of this study were to determine whether the extraction procedures previously used to obtain humin samples affect the chemical and physical structures of the organic carbon fraction in humins and to investigate their variation in sorption affinity and isotherm nonlinearity.

Materials and Methods Humin Extraction. Two peat soils were used in this study: Amherst peat soil (Terric Haplosaprist) was sampled from Lawrence swamp in Amherst, MA, and Florida peat (Pahokee muck) soil (Lithic Haplosaprists) was purchased from IHSS (International Humic Substances Society). Florida peat sample is from Palm Beach County, FL. Amherst peat soil developed on glaciolacustrine deposits with no specific clay mineral dominating (32). Florida peat soil usually formed in organic deposits of freshwater marshes over limestone. Soil residues were collected after six extractions of FA and HA with 0.1 M Na4P2O7 at a ratio of 1:10 (33). The four methods adopted for obtaining different humins were: (1) treating the soil residues with 1.0 M HF at a ratio of 1:10 for 24 h four times; (2) treating the soil residues with 6 M HF/HCl at 60 ( 2 °C at a ratio of 1:10 for 24 h; (3) repeatedly extracting the soil residues with 0.1 M NaOH at a ratio of 1:10 until no color remained in the supernatant (this approach is defined as the exhaustive extraction); and (4) deashing the soil residues with 1.0 M HF at a ratio of 1:10 for 24 h four times after the 0.1 M NaOH exhaustive extraction. We labeled these four extraction methods as HF, HF/HCl, NaOH, and NaOH-HF, respectively. Humin samples obtained from the four different extraction procedures were washed with deionized water, freeze-dried, and ground to pass through a 0.15 mm sieve. 8334

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 21, 2005

Solid-State 13C NMR Spectroscopy. Solid-state 13C NMR experiments were performed using a Bruker Avance DSX300 spectrometer at a 13C frequency of 75.48 MHz in a 7 mm zirconia rotor with a Kel-F cap. The 13C spectra were obtained using a total sideband suppression pulse program (CP/MASTOSS). The 90° 1H pulse length was 5 µs, the spinning speed was 5 kHz, the contact time was 1.5 ms, the acquisition time was 25 ms, and the recycle delay was 1 s. The total number of scans was around 7000, and 25 Hz line broadening was applied to the spectra. The structural carbons were interpreted as aliphatic carbons (0-50 ppm), carbohydrate carbons (50-108 ppm), aromatic carbons (108-162 ppm), and carboxylic/carbonyl carbons (162-220 ppm) (34). Proton spin-spin relaxation time (T2) was measured using the method described by Gunasekara et al. (35). Briefly, the pulse program used was the Carr-Purcell pulse sequence followed by Gill-Meibohm phase cycling. The humin sample was packed into a 7 mm zirconia rotor with Kel-F cap and analyzed in a static mode, because spinning-induced sidebands complicated the T2 analysis. The 1H T2 decay curve contained many data points for decreasing proton signal intensity with increasing time. Relaxation data were first fitted by an exponential decay function, which relates protons in a relatively mobile region

( )

t IE(t) ) IE(0) exp T2E

where IE(t) is the intensity for the exponential portion of the T2 decay curve at time t, IE(0) is the initial intensity of the exponential portion at time ) 0, t is time in µs, and T2E is the T2 (µs) value for the mobile domains. The Gaussian decay function, which relates the protons in relatively rigid regions, was used to fit data after subtraction of the exponential portion from the overall intensity data

[ ( )]

1 t IG(t) ) IG(0) exp 2 T2G

2

where IG(t), IG(0), t, and T2G are the parameters analogous to the exponential function, but for the protons in the rigid domains. The initial intensities for the exponential portion [IE(0)] and Gaussian portion [IG(0)] of the T2 decay curve were derived by fitting the exponential and Gaussian functions, respectively. Then the percent domain distributions were calculated as follows:

mobile domain (%) ) rigid domain (%) )

IE(0) IE(0) + IG(0) IG(0)

IG(0) + IE(0)

× 100

× 100

More detailed description and explanation on the methodology and data processing may be found in the paper by Gunasekara et al. (35). Sorption Experiments. Phenanthrene was used as a model HOC. The [ring-UL-14C] and unlabeled phenanthrene (>98% purity) were purchased from Sigma-Aldrich Chemical Co. (St. Louis, MO) and used without further purification. Sorption experiments were conducted using a batch equilibration method at 20 ( 1 °C. The sorption isotherms were obtained using 100-mL glass bottles with Teflon-lined septa cap. The background solution (pH 7.0) was 0.01 M CaCl2 with 200 µg/mL HgCl2 as a biocide. The 14C-labeled phenanthrene and its nonradioactive stock solutions were mixed with humin samples at different solid-to-solution ratios. The humin concentrations in the glass bottles were adjusted to achieve 30-70% sorption of phenanthrene. The suspensions were shaken for 3 days on hematology mixers to achieve

TABLE 1. Organic Carbon and Ash Contents of Humin Samples Obtained after Different Extraction Methodsa Amherst humin humin HF HF/HCl NaOH NaOH-HF

33.6 62.9 23.7 57.0

Results and Discussion

Florida humin

organic carbon (%) ash (%) 41.1 1.4 59.2 5.8

humin

organic carbon (%) ash (%)

HF HF/HCl NaOH NaOH-HF

54.7 56.6 21.1 48.4

The carbon-normalized distribution coefficients can be compared, as they have the same unit (7).

9.5 3.9 50.0 27.9

a HF ) treating the soil residue (crude humin) with 1.0 M HF for 24 h four times; HF/HCl ) treating the soil residue with 6 M HF/HCl at 60 ( 2 °C for 24 h; NaOH ) repeatedly extracting the soil residue with 0.1 M NaOH until no color remained in the supernatant (i.e., exhaustive extraction); and NaOH-HF ) after the 0.1 M NaOH exhaustive extraction, deashing the soil residue with 1.0 M HF for 24 h four times.

apparent equilibrium. Then, bottles were centrifuged at 1000g for 30 min, and 1 mL supernatant was sampled for liquid scintillation counting (Beckman LS 3801). The amount of phenanthrene sorbed was calculated by the mass difference. Sorption to the glass bottles was found to be negligible. All sorption data were fitted to the logarithmic form of the Freundlich equation:

log Cs ) log Kf + N log Ce where Cs is the total sorbed concentration (µg/g), Ce is the final solution phase concentration (µg/g), Kf is the Freundlich coefficient (µg1-N mLN g-1), and N is the exponent indicating isotherm linearity. Isotherms were plotted by log Cs vs log Ce, and log Kf and N were obtained from the intercept and slope of the plot, respectively. Because of isotherm nonlinearity (N < 1), the unit of Kf can vary from sample to sample; thus, comparison of Kf values cannot be made. Therefore, a single-point distribution coefficient (Kd ) Cs/Ce, mL g-1) at a given final solution concentration (Ce) was derived and the carbon-normalized distribution coefficient (KOC) was calculated by dividing Kd by the organic carbon content (fOC).

The organic carbon and ash contents of all the humin samples are listed in Table 1. It is obvious that humin samples differed greatly in organic carbon and ash contents. The NaOHextracted humin had significantly higher ash contents and lower carbon content than other extraction methods. Treating the soil residues (i.e., after six extractions with 0.1 M Na4P2O7) with 6 M HF/HCl at 60 °C for 24 h was much more effective in reducing ash content than treatment with 1 M HF for 24 h four times at room temperature. Thus, the HF/HClextracted humin had the highest carbon content and lowest ash content among all the humin samples. In addition, humin samples obtained from HF and NaOH-HF extractions also differed greatly in ash contents, even though both approaches included treatment four times with 1 M HF at room temperature. It is reasonable that silicate minerals could have more exposure to HF after NaOH exhaustive extraction, and the deashing with HF would be more efficient than that with HF alone. Thus, the HF-extracted Amherst humin had higher ash content than the NaOH-HF-extracted Amherst humin. However, the HF-extracted Florida humin had lower ash content than the NaOH-HF-extracted Florida humin. This might be due to the fact that the Florida peat soil contains minerals other than silicate minerals; thus, deashing with HF was less effective. However, the exact cause requires further investigation. NMR Spectroscopy. The solid-state 13C NMR spectra (Figure 1) and their integrations (Table 2) demonstrated that Florida humins had fewer aliphatic carbons but more aromatic carbons than Amherst humins for each extraction method, suggesting a quite different structural composition of humin between the two soils. For each soil, both NaOHand NaOH-HF-extracted humins had relatively higher aliphatic carbon contents but lower aromatic carbon contents than the HF-treated humin. It has been known that humin displays more aliphatic nature than HA or FA from the same

FIGURE 1. Solid-state 13C spectra of humin samples extracted from Amherst peat soil (A) and Florida peat soil (B) by different methods: (a) HF, (b) NaOH, (c) NaOH-HF, and (d) HF/HCl. VOL. 39, NO. 21, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

8335

TABLE 2. Organic Carbon Characteristics of Humin Samples As Determined by Solid-State 13C NMRa

sample

aliphatic carbohydrate aromatic carboxylic (0-50 (50-108 (108-162 (162-220 polarity ppm) ppm) ppm) ppm) index

HF HF/HCl NaOH NaOH-HF

35 43 39 44

Amherst Humin 32 24 15 30 29 22 30 18

HF HF/HCl NaOH NaOH-HF

19 30 26 27

Florida Humin 33 35 15 39 30 30 34 31

TABLE 3. Spin-Spin Relaxation Times (T2) and Distribution of Rigid and Mobile Domains humin HF extracted

9 12 10 8

0.47 0.35 0.45 0.43

13 16 14 8

0.54 0.41 0.52 0.49

HF ) treating the soil residue with 1.0 M HF for 24 h four times; HF/HCl ) treating the soil residue with 6 M HF/HCl at 60 ( 2 °C for 24 h; NaOH ) repeatedly extracting the soil residue with 0.1 M NaOH until no color remained in the supernatant (i.e., exhaustive extraction); and NaOH-HF ) after the 0.1 M NaOH exhaustive extraction, deashing the soil residue with 1.0 M HF for 24 h four times. a

source (7, 36). Because we collected the soil residues after only six extractions with 0.1 M Na4P2O7 at a ratio of 1:10, some HA and FA would still remain in the residues. The NaOH exhaustive extraction was able to remove HA and FA more completely; thus, it is not surprising that the humins from the exhaustive extraction were enriched in aliphatic carbons, consistent with our previous studies (5, 7). The 6 M HF/HCl treatment at 60 °C substantially reduced the carbohydrate carbon (50-108 ppm) content of the humins. As seen in Table 2, the HF/HCl-extracted humin contained only half of the carbohydrate carbon content of those extracted by other methods. Moreover, compared with the HF-extracted humin, the HF/HCl-extracted humin had more aliphatic and aromatic components. It should be noted that although both the NaOH-extracted humin and the HF/HCl-extracted humin had enriched aliphatic carbon content as compared with the HF-extracted humin, the causes were different. NaOH extraction removed most of the remaining HA and FA (if not all) from the samples, resulting in higher aliphatic carbon contents. On the contrary, the HF/HCl extraction did not remove the remaining HA and FA fractions; instead, this procedure eliminated a large portion of carbohydrate components. As a result, aliphatic and aromatic carbon contents in the HF/HCl-extracted humin

HF/HCl extracted

HF extracted HF/HCl extracted

T2 (µs)

% domain distribution

Florida Humin mobile semimobile rigid mobile semimobile rigid

37.0 9.0 3.8 37.6 9.0 3.7

33 58 9 40 54 6

Amherst Humin mobile semimobile rigid mobile semimobile rigid

57.8 9.0 3.3 60.2 11.0 3.4

15 75 11 17 73 11

domain

were proportionally increased due to the preferential removal of carbohydrate carbons. Also, the HF/HCl-extracted humin sample had relatively higher aromatic carbon content than the NaOH-extracted humins (Table 2). The polarity of SOM has been reported to be closely related to the sorption of organic contaminants (7, 18, 37). In our study, a polarity index was calculated by the following equation:

carbohydrate C (50-108 ppm) + phenolic C (145-162 ppm) + carboxylic C (162-220 ppm) polarity ) total organic C (0-220 ppm) The calculation is modified from that used by Kile et al. (18) because we included the phenolic groups (145-162 ppm) as polar carbons. The index scores suggest that the HF-extracted humin had highest polarity while the HF/HCl-extracted humin (at 60 °C) had the lowest polarity (Table 2). In addition, for a given extraction procedure, Florida humin had a slightly higher polarity than Amherst humin. In all the spectra (Figure 1), the characteristic doublets of relatively narrow and high poly(methylene) signals can be clearly distinguished. The peaks centered at 32.8 and 31 ppm are assigned to crystalline poly(methylene) and amorphous poly(methylene) carbons, respectively (13). Although the

FIGURE 2. Expanded solid-state 13C spectra of humins in the nonpolar aliphatic regions for (A) Amherst humin samples and (B) Florida humin samples. Humin was extracted by four methods: (a) HF, (b) NaOH, (c) NaOH-HF, and (d) HF/HCl. 8336

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 21, 2005

FIGURE 3. Comparison of expanded 13C NMR line shapes in nonpolar aliphatic regions for Amherst humin samples with different treatments: (a) 6 M HF/HCl at room temperature, (b) deionized water at 60 °C, and (c) 6 M HF/HCl at 60 °C. amorphous and crystalline poly(methylene) carbons have very similar chemical structures, they exhibit slightly different chemical shifts due to their different conformation. Figure 2 shows expanded views of the nonpolar aliphatic carbon NMR regions (0-50 ppm) of Amherst and Florida humins. The intensity changes of the doublets imply that amorphous poly(methylene) components had been augmented in the HF/HCl-extracted humin while crystalline components were the major fraction of poly(methylene) carbons in the humin samples by other extraction methods. Therefore, the 6 M HF/HCl treatment at 60 °C not only reduced carbohydrates dramatically but also altered the structural conformation of nonpolar aliphatic components by increasing the amount of amorphous domains. The amorphous nonpolar aliphatic domains were reported to exhibit rubbery-like mobility, which is favorable for sorption of nonpolar organic chemicals (38). It is understood that 1H spin-spin relaxation (T2) is related to molecular motion: protons in rigid regions relax rapidly,

represented by short T2 values, while the protons in mobile regions relax slowly, corresponding to long T2 values. In our study, the proton T2 relaxation of the HF-extracted humin and HF/HCl-extracted humin were measured and the relaxation data were processed on the basis of the method described by Gunasekara et al. (35). The only difference in data processing is that we fitted the exponential function twice instead of once, considering that there was no sharp boundary between rigid and mobile domains in reality, due to the heterogeneity of humic substances. We then defined the portion of the T2 decay obtained by the second exponential fit as semimobile domains, which describes the phase between rigid and mobile states. Table 3 summarized the T2 values and percent domain (proton population) distribution. Apparently, the HF/HCl-extracted Florida humin contained relatively more mobile domains than HF-extracted humin, which is in agreement with our 13C NMR results that HF/ HCl-extracted humin had more amorphous poly(methylene) domains. Amherst humin also had a similar trend, but with a slight difference in percent distribution. The increased mobile domains might be the consequence of deashing and reducing carbohydrate components. The treatment of 6 M HF/HCl at 60 °C was more effective than 1 M HF at room temperature to remove ash content (Table 1). The released humic materials due to removal of minerals were more likely to have higher mobility than those associated with minerals (14, 21, 22). On the other hand, reducing carbohydrate components might also loose the humin structure as a result of reduction in bond linkages (e.g., H-bonds) between different SOM structural moieties. Protons in mobile domains had longer T2 in Amherst humin than in Florida humin, but Florida humin had a higher percentage of mobile domains than Amherst humin, reflecting the structural and conformational differences between the two humins. However, the proton T2 method seems not a very sensitive measure here for monitoring distribution changes of mobile/condensed domains, especially for Amherst humin samples. This may be due to the fact that domains in Amherst humin are already very mobile, as indicated by the high sorption linearity described in the sorption section below and the large T2 (58 µs) in the mobile domain (Table 3). To discriminate the effect of temperature from that of 6 M HF/HCl, we treated the soil residues with 6 M HF/HCl at

FIGURE 4. Phenanthrene sorption isotherms of humin samples extracted by different methods from Florida peat soil. VOL. 39, NO. 21, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

8337

TABLE 4. Isotherm Parameters for Phenanthrene Sorption by Soil Humins Extracted by Different Methodsa concentration-dependent Koc (mL g-1)

N

log Kf

(µg1-N

mLN

g-1)

Ce ) 0.01 µg mL-1

Ce ) 0.1 µg mL-1

HF HF/HCl NaOH NaOH-HF Amherst humic acid

0.882 ( 0.007 0.894 ( 0.014 0.868 ( 0.009 0.910 ( 0.007 0.883 ( 0.011

Amherst Humin 3.982 ( 0.013 4.592 ( 0.030 3.841 ( 0.016 4.301 ( 0.013 3.868 ( 0.023

49 200 101 000 53 700 53 000 23 200

37 500 79 100 39 600 43 100 17 700

HF HF/HCl NaOH NaOH-HF Florida humic acid

0.648 ( 0.010 0.743 ( 0.009 0.553 ( 0.010 0.598 ( 0.008 0.893 ( 0.009

Florida Humin 3.945 ( 0.021 4.236 ( 0.019 3.527 ( 0.018 3.947 ( 0.017 3.869 ( 0.009

81 500 99 400 125 000 116 000 22 000

36 200 55 000 44 700 46 100 17 200

a HF ) treating the soil residue with 1.0 M HF for 24 h four times; HF/HCl ) treating the soil residue with 6 M HF/HCl at 60 ( 2 °C for 24 h; NaOH ) repeatedly extracting the soil residue with 0.1 M NaOH until no color remained in the supernatant (i.e., exhaustive extraction); and NaOH-HF ) after the 0.1 M NaOH exhaustive extraction, deashing the soil residue with 1.0 M HF for 24 h four times. Koc ) (Cs/Ce)/foc; Cs is the solid-phase equilibrium concentration, Ce is the liquid-phase equilibrium concentration, and foc is the organic carbon fraction of humin samples. Because of nonlinearity, isotherm slopes from direct linear fitting cannot be used for comparison and Kf values cannot be used either, as a result of different N among the samples (7, 43); thus, concentration-dependent Koc is calculated for the purpose of comparison.

FIGURE 5. Phenanthrene sorption isotherms of humin samples extracted by different methods from Amherst peat soil. the room temperature for 24 h and with deionized water at 60 °C for 24 h. The NMR spectra of Amherst humin are shown as an example in Figure 3. Higher temperature alone could increase the composition of amorphous poly(methylene), consistent with the data by Hu et al. (13), but the change was not as much as in the humin extracted by 6 M HF/HCl at 60 °C (Figure 3). The integration suggests that neither 6 M HF/HCl nor high temperature could reduce carbohydrate carbon content dramatically (data not shown), indicating that the increased amount of amorphous poly(methylene) and reduced carbohydrate content were the consequence of the combined action of 6 M HF/HCl and elevated temperature. Phenanthrene Sorption. The mineral phase is usually considered to be inert with respect to the sorption of HOCs. It is widely accepted that in soil/sediment-water systems where organic carbon content is >0.1%, sorption of HOCs in SOM predominates over adsorption by minerals, because of strong suppression by water of solute adsorption on polar mineral surfaces (39-41). Our previous study also demon8338

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 21, 2005

strated that humic acid could sorb phenanthrene 3 orders of magnitude more than clean minerals (montmorillonite and kaolinite) in terms of sorption coefficient (17). Therefore, the direct contribution of the mineral phase to the overall phenanthrene sorption is believed to be negligible and was not considered in the sorption calculation of humin samples; the lowest organic carbon content of these humin samples was 21% (Table 1). Nevertheless, minerals have been reported to affect sorption characteristics of SOM via changing the structural conformation of organic matter when SOM content is very low (1, 17). The alteration of organic carbon structural composition and conformation of humins impacted their sorption behaviors. For Florida humin, the NaOH exhaustively extracted humin had the lowest N value and the HF/HCl-extracted humin had the highest (Figure 4, Table 4). Deashing humin with HF after NaOH exhaustive extraction (NaOH-HF method) led to a higher N than that without deashing (NaOH method). The N values of humin samples obtained by using different extraction methods were well-correlated to their

ash contents. It has been considered that interaction of aliphatic components of organic matter with mineral surface might lead to rearrangement of amorphous domains into a condensed configuration, hence, resulting in more nonlinear isotherms (21). Deashing humin with HF or HF/HCl could release the mineral-associated humin components and reduce the contact between minerals and SOM; thus, the N value would be increased. Compared to 1 M HF, 6 M HF/HCl at 60 °C was more efficient in deashing, and more importantly, this extraction procedure produced a humin with an increased amount of amorphous nonpolar aliphatic components, which would explain the relatively higher N value of the HF/HCl-extracted humin than the HF-extracted humin. A similar trend was observed for Amherst humin (Figure 5), but the variation of N values was not as great as for Florida humin. According to the mobile/rigid concept of SOM and its relationship with sorption nonlinearity, Amherst humin was more mobile in structure than Florida humin, as represented by higher N values for each extraction method. This can also be confirmed by 13C NMR integration results (Table 2) and proton T2 values (Table 3). Proton T2 value of Amherst humin was much larger than that of Florida humin in the mobile domain. Florida humin had fewer aliphatic carbons and more aromatic carbons than Amherst humin, where aromatic components are usually considered to be more rigid than aliphatic moieties (10, 11). However, Amherst humin contained similar or even more ash content than Florida humin; for instance, the HF- and NaOHextracted Amherst humins contained 41% and 59% ash, respectively, while the HF- and NaOH-extracted Florida humin had 10% and 50% ash content, respectively. Thus, sorption nonlinearity and its variation cannot be simply interpreted by ash content alone. The chemical composition and structure of organic matter in humin seem the primary contributing factors. Moreover, mineral compositions in soils and their interactions with organic matter might also contribute to the sorption nonlinearity and its fluctuation (17), which needs further examination using actual soil and humin samples. Large differences in KOC were observed for the humin samples extracted by different methods (Table 4). For Amherst humin, the KOC value of the HF/HCl-extracted humin was nearly twice as high as those of humins extracted by other methods. The sorption affinity order in terms of KOC value by extraction methods was HF < NaOH ≈ NaOH-HF < HF/HCl at both equilibrium concentrations of 0.01 and 0.1 µg mL-1. The same trend was observed for Florida humin at Ce ) 0.1 µg mL-1; however, at the concentration of 0.01 µg mL-1 the HF/HCl-extracted humin had slightly lower KOC than the NaOH- and NaOH-HF-extracted humins due to isotherm nonlinearity. Regardless of extraction methods, both Amherst and Florida humins had substantially higher KOC than their corresponding HAs (Table 4). Because the HF-extracted humin contained a certain amount of HA, removal of such a HA fraction by NaOH exhaustive extraction corresponded to the increased amount of aliphatic components, resulting in increased KOC values. As for the HF/ HCl-extracted humin, the enhancement of phenanthrene sorption might be attributed to an increased amount of amorphous poly(methylene) components and reduced polarity due to the significant reduction of carbohydrate components (42). It has been reported that the fraction of amorphous poly(methylene) has a strong positive correlation with the sorption of nonpolar molecules (38), and the low polarity of SOM is favorable for HOCs sorption (7, 18). In summary, humin samples obtained from various extraction methods differed greatly in their chemical and structural composition, ash content, and sorption properties of HOCs. The HF extraction is the simplest method to prepare

a humin sample; however, the sample from this approach is really a mixture of humin, HA, and FA. As compared to HF extraction, the 0.1 NaOH exhaustive extraction successfully removed HA from soil samples, resulting in more aliphatic components (0-50 ppm) and consequently higher KOC values. We believe that samples prepared by NaOH exhaustive extraction represent the characteristics of humin. The deashing with HF after NaOH exhaustive extraction effectively lowered ash contents of humin samples with only slight changes to their chemical structure and sorption properties. Therefore, both NaOH and NaOH-HF would be appropriate approaches for humin extraction. The extraction with 6 M HF/HCl at 60 °C not only removed ash content efficiently but also substantially modified the chemical structure of humin, as indicated by the dramatically reduced carbohydrate components (50-108 ppm). Further, this 6 M HF/HCl extraction also enhanced the amount of amorphous poly(methylene) domains. Therefore, this extraction method would be discouraged for use due to structural modifications of humin both chemically and physically.

Acknowledgments This work was supported in part by the CSREES, USDA National Research Initiative Competitive Grants Program (2005-35107-15278), Research Grant No. IS-3385-03R from BARD, the United States-Israel Binational Agricultural Research and Development Fund, and the Federal Hatch Program (Project No. MAS00860).

Literature Cited (1) Salloum, M. J.; Dudas, M. J.; McGill, W. B. Variation of 1-naphthnol sorption with organic matter fractionation: The role of physical conformation. Org. Geochem. 2001, 32, 709719. (2) Borisover, M.; Graber, E. R. Hydration of natural organic matter: Effect on sorption of organic compounds by humin and humic fractions vs original peat material. Environ. Sci. Technol. 2004, 38, 4120-4129. (3) Rice, J. A. Humin. Soil Sci. 2001, 166, 848-857. (4) Stevenson, F. J. Humus Chemistry: Genesis, Composition, Reactions, 2nd ed.; Wiley: New York, 1994. (5) Xing, B.; Liu, J. D.; Liu, X. B.; Han, X. Z. Extraction and characterization of humic acids and humins from a black soil of China. Pedosphere 2005, 15, 1-8. (6) Kohl, S. D.; Rice, J. A. The binding of contaminants to humin: A mass balance. Chemosphere 1998, 36, 251-261. (7) Kang, S.; Xing, B. Phenanthrene sorption to sequentially extracted soil humic acids and humins. Environ. Sci. Technol. 2005, 39, 134-140. (8) Preston, C. M.; Newman, R. H. Demonstration of spatial heterogeneity in the organic matter of deashed humin samples by solid-state 13C CP-MAS NMR. Can. J. Soil Sci. 1992, 72, 1319. (9) Perminova, I. V.; Grechishcheva, N. Y.; Kovalevskii, D. V.; Kudryavtsev, A. V.; Petrosyan, V. S.; Matorin, D. N. Quantification and prediction of the detoxifying properties of humic substances related to their chemical binding to polycyclic aromatic hydrocarbons. Environ. Sci. Technol. 2001, 35, 3841-3848. (10) Xing, B.; Chen Z. Spectroscopic evidence for condensed domains in soil organic matter. Soil Sci. 1999, 164, 40-47. (11) Johnson, M. D.; Huang, M.; Weber, W. J. A distributed reactivity model for sorption of soils and sediments. 13. Simulated diagenesis of natural sediment organic matter and its impact on sorption/desorption equilibria. Environ. Sci. Technol. 2001, 35, 1680-1687. (12) Kogel-Knabner, I.; Hatcher, P. G.; Tegelaar, E. W.; de Leeuw, J. W. Aliphatic components of forest soil organic matter as determined by solid-state 13C NMR and analytical pyrolysis. Sci. Total Environ. 1992, 113, 89-106. (13) Hu, W.-G.; Mao, J.-D.; Xing, B.; Schmidt-Rohr, K. Poly(methylene) crystallites in humic substances detected by nuclear magnetic resonance. Environ. Sci. Technol. 2000, 34, 530-534. (14) Cuypers, C.; Grotenhuis, T.; Nierop, K. G. J.; Franco, E. M.; de Jager, A.; Rulkens, W. Amorphous and condensed organic matter domains: The effect of persulfate oxidation on the composition VOL. 39, NO. 21, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

8339

(15)

(16) (17) (18)

(19) (20) (21) (22)

(23) (24)

(25) (26)

(27)

(28)

(29)

of soil/sediment organic matter. Chemosphere 2002, 48, 919931. Huang, W.; Weber, W. J., Jr. A distributed reactivity model for sorption by soils and sediments. 10. Relationships between desorption, hystersis, and the chemical characteristics of organic domains. Environ. Sci. Technol. 1997, 31, 2562-2569. Kohl, S. D.; Rice, J. A. Contribution of lipids to the nonlinear sorption of polycyclic aromatic hydrocarbons to soil organic matter. Org. Geochem. 1999, 30, 929-936. Wang, K.; Xing, B. Structural and sorption characteristics of adsorbed humic acid on clay minerals. J. Environ. Qual. 2005, 34, 342-349. Kile, D. E.; Wershaw, R. L.; Chiou, C. T. Correlation of soil and sediment organic matter polarity to aqueous sorption of nonionic compounds. Environ. Sci. Technol. 1999, 33, 20532056. Xing, B. Sorption of anthropogenic organic compounds by soil organic matter: A mechanistic consideration. Can. J. Soil Sci. 2001, 81, 317-323. Xing, B.; Pignatello, J. J. Dual-mode sorption of low polarity compounds in glassy polyvinychloride and soil organic matter. Environ. Sci. Technol. 1997, 31, 792-799. Gunasekara, A. S.; Xing, B. Sorption and desorption of naphthalene by soil organic matter: Importance of aromatic and aliphatic components. J. Environ. Qual. 2003, 32, 240-246. Watanabe, N.; Schwartz, E.; Scow, K. M.; Young, T. M. Relating desorption and biodegradation of phenanthrene to SOM structure characterized by quantitative pyrolysis GC-MS. Environ. Sci. Technol. 2005, 39, 6170-6181. Ding, G.; Novak, J. M.; Herbert, S.; Xing, B. Long-term tillage effects on soil metolachlor sorption and desorption behavior. Chemosphere 2002, 48, 897-904. Dec, J.; Haider, K.; Schaffer, A.; Fernandes, E.; Bollag, J. M. Use of a silyation procedures and 13C NMR spectroscopy to characterize bound and sequestered residues of cyprodinil in soil. Environ. Sci. Technol. 1997, 31, 2991-2997. Nam, K.; Kim, J. Y. Role of loosely bound humic substances and humin in the bioavailability of phenanthrene aged in soil. Environ. Pollut. 2002, 118, 427-433. Tremblay, L.; Kohl, S. C.; Rice, J. A.; Gagne, J. P. Effects of lipids on the sorption of hydrophobic organic compounds on geosorbents: A case study using phenanthrene. Chemosphere 2005, 58, 1609-1620. Guignard, C.; Lemee, L.; Ambles, A. Lipid constituents of peat humic acids and humin. Distinction from directly extractable bitumen components using TMAH and TEAAc thermochemolysis. Org. Geochem. 2005, 36, 287-297. Chiou, C. T.; Kile, D. E.; Rutherford, D. W.; Sheng, G.; Boyd, S. A. Sorption of selected organic compounds from water to a peat soil and its humic acid and humin fractions: Potential sources of the sorption nonlinearity. Environ. Sci. Technol. 2000, 34, 1254-1258. Smernik R. J.; Oades, J. M. Spin accounting and RESTOREs Two new methods to improve quantitiation in solid-state 13C NMR analysis of soil organic matter. Eur. J. Soil Sci. 2003, 54, 103-116.

8340

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 21, 2005

(30) Song, J.; Peng, P.; Huang, W. Black carbon and kerogen in soils and sediments. 1. Quantification and characterization. Environ. Sci. Technol. 2002, 36, 3960-3967. (31) Jeong, S.; Werth, C. J. Evaluation of methods to obtain geosorbent fractions enriched in carbonaceous materials that affect hydrophobic organic chemical sorption. Environ. Sci. Technol. 2005, 39, 3279-3288. (32) Tiner, R. W.; Veneman, P. L. M. Hydric Soils of New England; University of Massachusetts cooperative extension, Bulletin C-183, Amherst, MA. 1987. (33) Swift, R. Chemical methods. In Methods of Soil Analysis; Sparks, D. L., Ed.; Soil Science Society of America: Madison, WI, 1996; Part 3. (34) Mao, J.-D.; Hu, W.-G.; Schmidt-Rohr, K.; Davies, G.; Ghabbour, E. A.; Xing, B. Quantitative characterization of humic substances by solid-state carbon-13 nuclear magnetic resonance. Soil Sci. Soc. Am. J. 2000, 64, 873-884. (35) Gunasekara, A. S.; Simpson, M. J.; Xing, B. Identification and characterization of sorption domains in soil organic matter using structurally modified humic acids. Environ. Sci. Technol. 2003, 37, 852-858. (36) Lichtfouse, E.; Chenu, C.; Baudin, F.; Leblond, C.; Silva, M. D.; Behar, F.; Derenne S.; Largeau, C.; Wehrung, P.; Albrecht, P. A novel pathway of soil organic matter formation by selective preservation of resistant straight-chain biopolymers: Chemical and isotope evidence. Org. Geochem. 1998, 28, 411-415. (37) Xing, B. The effect of the quality of soil organic matter on sorption of naphthalene. Chemosphere 1997, 35, 633-642. (38) Mao, J.-D.; Hundal, L. S.; Thompson, M. L.; Schmidt-Rohr, K. Correlation of poly(methylene)-rich amorphous aliphatic domains in humic substances with sorption of a nonpolar organic contaminant, phenanthrene. Environ. Sci. Technol. 2002, 36, 929-936. (39) Schwarzenbach, R. P.; Westall, J. Transport of nonpolar organic compounds from surface water to groundwater: Laboratory sorption studies. Environ. Sci. Technol. 1981, 15, 1360-1367. (40) Kile, D. E.; Chiou, C. T.; Zhou, H.; Li, H.; Xu, O. Partition of nonpolar organic pollutants from water to soil and sediment organic matters. Environ. Sci. Technol. 1995, 29, 1401-1406. (41) Yang, K.; Zhu, L.; Lou, B.; Chen, B. Significance of natural organic matter in nonlinear sorption of 2,4-dichlorophenol onto soils/ sediments. Water Resour. Res. 2004, 40, W01603. (42) Chen, B.; Johnson, E. J.; Chefetz, B.; Zhu, L.; Xing, B. Sorption of polar and nonpolar aromatic organic contaminants by plant cuticular materials: The role of polarity and accessibility. Environ. Sci. Technol. 2005 39, 6138-6146. (43) Carmo, A. M.; Hundal, L. S.; Thompson, M. L. Sorption of hydrophobic organic compounds by soil materials: Application of unit equivalent Freundlich coefficients. Environ. Sci. Technol. 2000, 34, 4363-4369.

Received for review April 15, 2005. Revised manuscript received August 6, 2005. Accepted August 29, 2005. ES050737U