JEQ Journal of Natural Resources and Life Sciences Education
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (6)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Eriksson, J.
Right arrow Articles by Skyllberg, U.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Eriksson, J.
Right arrow Articles by Skyllberg, U.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by Eriksson, J.
Right arrow Articles by Skyllberg, U.
Related Collections
Right arrow Organic Compounds
Right arrow Soil Pollution
Right arrow Soil Chemistry
Right arrow Forest Soils
Journal of Environmental Quality 30:2053-2061 (2001)
© 2001 American Society of Agronomy, Crop Science Society of America, and Soil Science Society of America

TECHNICAL REPORT
Organic Compounds in the Environment

Binding of 2,4,6-Trinitrotoluene and its Degradation Products in a Soil Organic Matter Two-Phase System

J. Eriksson*,a,b and U. Skyllberga

a Dep. of Forest Ecology, Swedish Univ. of Agriculture, SLU, S-901 83 Umeå, Sweden
b Defence Research Establishment, FOA, S-901 82 Umeå, Sweden

* Corresponding author (Johan.Eriksson{at}sek.slu.se)

Received for publication January 22, 2001.

   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
The widely used explosive 2,4,6-trinitrotoluene (TNT) and its degradation products are of large environmental concern because of their toxic properties and high concentrations encountered in contaminated soils. Batch experiments were used to study TNT* (the sum of TNT and its degradation products) bonding to dissolved (DOM) and particulate (POM) soil organic matter. Reversed-phase high performance liquid chromatography (RP-HPLC) was used as a separation technique in combination with 14C-labeled TNT to determine free TNT and TNT* bound to DOM. By use of dialysis we showed that DOM did not interfere with the HPLC analysis of free TNT. Depending on pH and total TNT concentration, the relative distribution of TNT* among water, POM, and DOM varied between 60 to 90, 10 to 30, and 0.5 to 6%, respectively, after 22 h of equilibration. The association of TNT* to DOM was strongly pH dependent and followed a nonlinear Langmuir isotherm. The association of TNT* to POM was less pH dependent and data were equally well fitted by linear and nonlinear isotherms. Particulate organic matter had 6.4 (pH 6.2) to 22 (pH 5.2) times greater capacity to bind TNT* than DOM, but the binding strength (the slope of the isotherm) was greater for DOM. The TNT degradation was enhanced with increasing concentration of soil organic matter, resulting in a stronger bonding of TNT* to DOM and POM. Based on our results, combined with other recent findings, we suggest that it is mainly the degradation products of TNT that associate with DOM and POM, and that the association with DOM is mainly of ionic character involving specific DOM sites. The greater binding capacity and a weaker, linear type of isotherm suggests a nonspecific type of partitioning in POM, possibly of hydrophobic character.

Abbreviations: ADNT, aminodinitrotoluene • DANT, diaminonitrotoluene • DOM, dissolved soil organic matter • FA, fulvic acid • HA, humic acid • HPLC, high performance liquid chromatography • POM, particulate soil organic matter • RP, reversed phase • SOM, soil organic matter • TNT, trinitrotoluene • TNT*, sum of trinitrotoluene and its degradation products


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
TRINITROTOLUENE, a widely used explosive, and its degradation products have toxic and mutagenic effects on several organisms, including humans (Kaplan and Kaplan, 1982). In Sweden, ammunition found in areas used for incineration and dumping is considered an environmental threat and since 1995 has been the subject of investigation by the Defence Research Establishment (FOA). Risk assessments are focused on the transportation of TNT and its degradation products in soils and how these products may reach and affect biological systems. In order to understand how these substances are retained or transported in soils we need to study in detail their affinity for mobile and nonmobile soil constituents. Of special importance is the bonding of TNT and its degradation products to dissolved (mobile) and particulate (mostly nonmobile) soil organic matter.

Trinitrotolune degrades by a stepwise reduction of its three nitro groups to corresponding amino groups. The first nitro group is reduced even during aerobic conditions while reduction of the last nitro group demands strictly anaerobic conditions (McCormick et al., 1976; Rieger and Knackmuss, 1995). Intermediates in this degradation pathway, such as nitroso and hydroxylamino derivatives and the azoxy products, are very reactive and have recently been shown to bind to soil organic matter (SOM) within hours or days (Achtnich et al., 1999a). Trinitrotoluene, as well as other nitroaromatic compounds (NAC), are also known to bind to negatively charged 2:1 phyllosilicates (Haderlein et al., 1996), possibly via an electron donor–acceptor complex (EDA) involving an electron excessive clay surface and the electron deficient NAC (Haderlein and Schwarzenbach, 1993; Weissmahr et al., 1997). Thus, retention and mobility of TNT and its degradation products in soil are highly dependent on the availability of 2:1 phyllosilicates and SOM. Water-soluble organic matter is of special interest if it associates with TNT and its degradation products, possibly enhancing their mobility.

Soil organic matter may be separated into dissolved (DOM) and particulate (POM) organic matter. Dissolved organic matter comprises a complex mixture of organic substances with sizes between 200 and 100000 Da. A variable but dominant portion of the organic substances in soil can be described as humic substances (i.e., undefined macromolecules largely built up of aromatic and aliphatic moieties) (Stevenson, 1994). An obvious difference between DOM and POM is the size of the organic molecules, being smaller for the former. Otherwise, little is known about chemical and structural differences between DOM and POM. Important for the bonding of organic chemicals to SOM is the composition and density of reactive organic groups and hydrophobic moieties in DOM and POM (Engebretson and von Wandruszka, 1997, 1998). The strong tendency of nonpolar organic chemicals to associate with SOM through hydrophobic partition is well known (Hassett et al., 1983; Chiou, 1989), whereas possible specific bonding involving functional groups of SOM is less studied and not well understood.

The limited knowledge about the binding of TNT and its degradation products to SOM is largely based on studies of extracted humic acids (HA). Li et al. (1997) reported a strong binding of TNT, aminodinitrotoluene (ADNT), and diaminonitrotoluene (DANT) to Aldrich (Milwaukee, WI) humic acid after 48 h of equilibration during nonsterile conditions, while Held et al. (1997) found no significant association of TNT to HA under sterile conditions. This indicates that it is the reduced degradation products and not TNT that form bonds with SOM. This was recently verified using 15N-labeled TNT (Achtnich et al., 1999a; Knicker et al., 1999; Bruns-Nagel et al., 2000). Results based on 15N nuclear magnetic resonance (NMR) suggest that hydroxylamino and azoxy compounds initially (within 4 d after TNT addition) associate with organic functional groups through ionic interactions. With increasing incubation time up to 25 d and non-oxidizing conditions, a covalent association between SOM and the amino derivatives ADNT and DANT is formed (Achtnich et al., 1999a). Interesting to note is that soluble and potentially soluble organic substances (HA and fulvic acid [FA]) seem to have a limited and substantially lower capacity to bind TNT degradation products as compared with the nonsoluble humin fraction (Drzyzga et al., 1998; Achtnich et al., 1999a). This might indicate different bonding characteristics for TNT* to DOM and POM due to chemical and physical differences between these two fractions of SOM.

In this study the binding of TNT* to DOM and POM is determined simultaneously, in what we designate the soil organic matter two-phase system. Concentrations of DOM in relation to POM were chosen to be representative for temperate acidic forest soils. We also investigate possible interference of DOM on the determination of free TNT and its degradation products by HPLC, something that, to our knowledge, has not been considered in earlier studies.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Organic Soil, Dissolved Organic Matter, and TNT
Soil organic matter was collected from the organic horizon of a Spodosol (Soil Survey Staff, 1992), located near Umeå in northern Sweden. The upper 2 cm of the organic horizon (the less-decomposed fermentation layer) was sampled and homogenized through a 4-mm cutting sieve. The soil material was stored in darkness at 4°C. The concentration of SOM was determined as total organic carbon by dry combustion (2400 CHN elemental analyzer; PerkinElmer, Norwalk, CT). General soil chemistry data for the soil are presented in Table 1. Metal ions (Na+, K+, Mg2+, Ca2+, Mn2+, Al3+, and Fe3+) adsorbed to the soil were sequentially extracted with 0.5 M CuCl2 using the method of Skyllberg and Borggaard (1998). The total cation exchange capacity (CECt) was calculated as the sum of total acidity determined at pH 8.2 (Thomas, 1982) and charges pertaining to non-acidic cations (K+, Na+, Ca2+, and Mg2+). Water and inorganic matter content were measured as loss of weight after drying (105°C for 24 h) and ash content after ignition at (540°C for 6 h), respectively.


View this table:
[in this window]
[in a new window]
  		Table 1. Selected chemistry data of the organic soil.

 
Dissolved organic matter used in the sorption experiments (see below) was extracted from the organic soil by a slightly modified version of the method by Adams and Byrne (1989). Thirty grams of moist soil and 200 mL of Millipore (Bedford, MA) water was added to two 250-mL polycarbonate centrifuge bottles (Nalgene, Rochester, NY). To each bottle, 2.4 g of Na+–saturated Chelex 20 resin (Bio-Rad, Hercules, CA) was added to complex and remove Al3+, Ca2+, and other polyvalent cations adsorbed to SOM. After gentle shaking for 48 h, the bottles were centrifuged for 10 min at 6000 rpm on a J2-21M/E centrifuge (Beckman, Glenrothes, Scotland). The supernatant, which had a pH around 6, was decanted. The complete procedure was repeated once more. The combined DOM extract (stock solution), containing approximately 700 mg C L-1, was kept cool in the dark until use. The soil material with the resin was discarded.

The 14C-labeled TNT (22 Bq µg-1) was synthesized from 14C-labeled toluene at the Swedish Defence Research Establishment (Weapon and Protections Division). The TNT (approximately 100 mg L-1) was dissolved in water during stirring and careful heating. After the TNT was dissolved it was placed to cool at 4°C for 24 h. Before use, the TNT solution was filtered through a 0.45-µm polypropylene filter for subsequent analysis by HPLC. The TNT derivatives 2-ADNT, 4-ADNT, and 2,4-DANT were purchased from Promochem (Wesel, Germany). All chemicals were of analytical purity grade.

Sorption Experiments
Sorption experiments were conducted in DOM (one-phase) and DOM + POM (two-phase) systems at pH (3.7–6.2), ionic strength (1 and 50 mM NaCl), and concentrations of DOM (75–100 mg C L-1) relevant for natural conditions in organic-rich acidic forest soils. For the DOM + POM system the soil to solution ratio was kept low enough to obtain solution for all chemical analyses. This was achieved by adding DOM stock solution and Millipore water to a sample from the original organic soil (POM). The DOM + POM system was allowed to equilibrate for 24 h until 14C-TNT was added. Prior to each sorption experiment, the DOM stock solution was filtered through a 0.45-µm polypropylene filter and its absorbance at 254 nm was determined. The concentration of dissolved organic carbon (DOC) was calculated from the absorbance of known standards. A 0.45-µm filter was used to separate DOM and POM, and TNT* associated to each of these fractions in the DOM + POM system.

Experiments with only DOM are henceforth described as one organic phase (one-phase) systems, and experiments with DOM + POM are described as two organic phase (two-phase) systems. Thus, in the two-phase system SOM is the sum of POM and DOM. Note that concentrations of POM and DOM were determined as mass of organic carbon per liter solution.

One-Phase System, Dissolved Organic Matter
Trinitrotoluene and DOM were mixed in 50-mL Erlenmeyer flasks of borosilicate glass. In all experiments the flasks were shaken on a reciprocal shaker for 15 min (150 rpm) every second hour, even though the total time of equilibration varied. In all experiments a control, without organic matter but otherwise treated the same as the DOM solutions, was sampled for monitoring unwanted side reactions (e.g., adsorption of TNT to glass walls). Diluted solutions of HCl and Ca(OH)2 were used to adjust pH. All experiments were carried out in darkness and the temperature was kept at 22 ± 1°C. A constant ionic strength was maintained using NaCl.

Kinetic Experiments
The time dependence of sorption and the rate of formation of amino derivatives were investigated at two levels of ionic strength (1 and 50 mM of NaCl) and two levels of pH (3.7 and 6.0). The initial concentrations of TNT and DOM were 19 µM and 90 mg C L-1, respectively. The experimental solution was sampled with a syringe, filtered through 0.45 µm and analyzed after 1, 7, and 15 d.

Equilibrium Experiments
Initial concentrations of TNT ranged from 0.2 to 290 µM. Sorption was studied at pH 4.6 and 5.9. As ionic medium, 50 mM NaCl was used. The DOM concentration was 97 to 99 mg C L-1 and the equilibration time was 22 h.

Dialysis Experiments
Spectra/Por-3 dialysis tubing (Spectrum, Los Angeles, CA) with a molecular weight cut-off (MWCO) of 3500 containing DOM was placed in a glass beaker with a total volume of 270 mL of 1 mM NaCl solution. Initial total concentrations in the system (pH 6) were DOM = 89 mg C L-1 and TNT = 18 µM. Note that DOM was added and kept inside the dialysis membrane, whereas TNT was added outside of the membrane. The experimental solution was continuously stirred. The kinetics of the concentration of free TNT, bound TNT* to DOM, and degradation products outside and inside the membrane was studied by sampling 1 mL on both sides of the membrane with a syringe at six occasions during 13 d. In a control experiment without DOM, but otherwise treated the same way, other possible reactions aside from the TNT–DOM interaction were monitored.

Two-Phase System, Dissolved Organic Matter + Particulate Organic Matter
In Erlenmeyer flasks of borosilicate glass, organic soil (POM), DOM, TNT, Millipore water, and NaCl were mixed. The concentrations of free TNT and degradation products in the solution and bound to DOM were determined with HPLC and 14C activity, respectively. The concentration of TNT and its degradation products sorbed to POM was calculated by difference.

Kinetic Experiments
The kinetics in the two-phase system were studied at three POM concentrations: 0, 165, and 645 mg C L-1. The initial DOM concentration was 75 mg C L-1 and the initial concentration of TNT was 46 µM. The ionic strength was kept constant with NaCl at 50 mM. The pH was not adjusted but monitored. Aliquots of the suspension were sampled after 0.01, 0.08, 0.2, 1, and 7 d.

Equilibrium Experiments
The initial concentrations of TNT ranged from 0.1 to 300 µM. The concentration of POM was 650 mg C L-1, and the concentration of DOM was 100 mg C L-1. The ionic strength was kept constant with NaCl at 50 mM, and pH was adjusted to 4.4, 5.2, 5.6, and 6.2. The equilibration time was 20 h.

Chemical Analyses
TNT, Degradation Products, and Dissolved Organic Matter
Aliquots sampled in the one- and two-phase systems were filtered through a 0.45-µm polypropylene filter and analyzed with RP-HPLC without storage delay to prevent unwanted side reactions. The first 5 mL of the filtrate was discarded to avoid reactions with the filter and the following 1 mL was collected and analyzed. There was no significant difference in analyses results between filtered and nonfiltered samples (data not shown). The TNT, ADNT, and DANT peaks were identified by retention times combined with the UV spectra from known standards. Analysis was performed with a system of Waters (Milford, MA) autosampler 717, HPLC pump 510, and a photodiode array detector PAD 996. The hardware was controlled and monitored from a computer with the Millennium 32 software (Waters Corporation, 1998). Concentrations of free TNT, degradation products, and DOM were determined by absorption at 254 nm after separation. The column used for separation was a Lichrocart, 125-4, packed with 5-µm spheres of Puraspher, RP-18e (Merck, Darmstadt, Germany). The mobile phase was methanol (HPLC grade) and 10 mM phosphate buffer, with a pH of 7, in a proportion of 50:50 (v/v), but 30:70 when analyzing DANT. The flow rate was 1 mL min-1 and the injection volume was between 10 and 250 µL.

Carbon-14 Determination
Trinitrotoluene labeled with 14C was used to determine TNT associated with DOM. The HPLC analysis effluent was fractionated, collected, and subsequently analyzed with respect to 14C on a liquid scintillation system, LS 5000CE (Beckman, Fullerton, CA). The first 15 mL of one HPLC run was separated into five aliquots (Fig. 1) . The first fraction contained DOM and TNT* bound to DOM. The second and fifth fractions represented unknown activity of 14C, and the third and fourth fractions contained TNT and ADNT, respectively. Each fraction was mixed with Beckman scintillation cocktail before analyses. The 14C radioactivity associated with POM was calculated as the difference between total 14C radioactivity and the sum of radioactivity associated with the five HPLC fractions.



View larger version (33K):
[in this window]
[in a new window]
  		Fig. 1. Measured 14C activity inside and outside the dialysis membrane after 48 h of equilibration. A schematic high performance liquid chromatography with ultraviolet detection (HPLC–UV) chromatogram (254 nm, total run of 15 min) is presented in relation to the five collected fractions used in the 14C analysis. The first fraction contains the sum of trinitrotoluene and its degradation products (TNT*) bound to dissolved organic matter (DOM), the second and fifth fractions are unknown 14C, and the third and fourth contain TNT and aminodinitrotoluene (ADNT), respectively.

 
Sorption Isotherms
Sorption data were modeled using linear (Eq. [1]), Freundlich (Eq. [2]), and Langmuir (Eq. [3]) adsorption isotherms. Freundlich and Langmuir isotherms were fitted using nonlinear regression by minimizing the sum of squared differences between observed and fitted values. The computer software SigmaPlot (SPSS, 2000) (Marquardt–Levenberg algorithm) was used for this purpose:

In Eq. [1], [2], and [3], Cs represents the sorbed concentration of TNT* expressed in relation to the mass of organic carbon (mol kg-1 C). The molecular mass of TNT (227 g mol-1) was taken as an estimate of the molecular mass of bound TNT*. The term Cw is the equilibrium concentration of free TNT (mol L-1) in solution. The terms Koc and Kf are partitioning coefficients for the linear and Freundlich equation respectively (L kg-1 C) and N is the power of Cw in the Freundlich equation. In the Langmuir equation qmax is the maximum sorption capacity, assuming that the sorbate is arranged in a monolayer (µmol g-1 C) at the adsorbing surface and KL is the Langmuir constant. As will be shown below, TNT decomposes rapidly in soils and its decomposition products are believed to be involved in the sorption mechanisms (Achtnich et al., 1999a). Since we could not determine different forms of TNT derivatives adsorbed to POM, the isotherms above were used to describe the association of TNT* to DOM and POM. Thus, the fit of the adsorption isotherms to data gives us information about the adsorptive properties of DOM and POM but no information on the forms of TNT derivatives involved.


   RESULTS AND DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
CONCLUSIONS
 REFERENCES
 
Determination of Free TNT in Solution
A dialysis experiment, with 14C labeled TNT, was conducted in order to investigate possible interference from DOM on free TNT determination by HPLC. In Fig. 1 the 14C activity of the five HPLC fractions and a schematic HPLC–UV chromatogram are shown. There was a significant 14C activity in the first HPLC fraction. Since almost all of the activity was found inside the dialysis membrane, it is clear that this activity was associated with TNT* bound to DOM with a molecular size larger than 3500 Da. The low activity found outside the membrane was apparently associated with TNT* bound to organic molecules with a size less than 3500 Da or possibly with some unidentified polar degradation product, as reported by Achtnich et al. (1999b). Most of 14C activity was found in the third HPLC fraction, and by a comparison with standards we know that this activity includes free TNT in solution. The fact that similar activities were found outside and inside the membrane suggests that this fraction contains almost exclusively free TNT in chemical equilibrium with TNT bound to DOM. If the third fraction contained some TNT associated with DOM, the 14C activity would have been higher inside, as compared with outside of the membrane.

An interesting side effect was an enhanced degradation of TNT in the dialysis experiments. After 48 h approximately twice the amount of ADNT was found in dialysis experiments as compared with control experiments without dialysis tubing. This may have been caused by the high local concentration of DOM (and its associated microorganisms) inside the dialysis membrane, possibly enhancing the TNT degradation. However, also in the control with no DOM there was some degradation of TNT to ADNT, not seen in experiments without dialysis tubing. Thus, the dialysis tubing itself, made of regenerated cellulose, may have contributed to degradation of TNT by a catalysis process.

Kinetic Experiments
In the DOM one-phase system, free TNT in solution decreased rapidly within 24 h. At that time there was no measurable increase in ADNT (Fig. 2) . This indicates that the initial rapid decrease in free TNT was probably due to sorption of TNT* to DOM. After approximately 2 d TNT showed a steady, more linear decrease with time inversely related to an increase in the free concentration of ADNT. This indicates that decomposition of TNT probably was the most important process behind the decreasing concentration of TNT in solution. It can be noted that the ionic strength had little effect on sorption of TNT whereas pH had a substantial effect. Both the decomposition of TNT to ADNT and the binding of TNT* to DOM increased greatly with increasing pH (Fig. 2).



View larger version (28K):
[in this window]
[in a new window]
  		Fig. 2. Free trinitrotoluene (TNT) and aminodinitrotoluene (ADNT) concentrations as a function of time, pH, and ionic strength in the presence of 0 or 90 mg C L-1 of dissolved organic matter (DOM) (one-phase system). Filled symbols refer to pH 6.0 and open to pH 3.7. The initial concentrations were calculated. IS = ionic strength.

 
Also in the DOM + POM two-phase system the free concentration of TNT decreased with time (Fig. 3) . The decrease can be divided into two phases: an inverse exponential phase grading into a linear phase after less than 24 h. This pattern was most obvious at the two highest concentrations of DOM + POM (240 and 720 mg C L-1). Free ADNT was detected after 24 h (1% of added TNT) and increased to 10% of added TNT after 7 d for the highest DOM + POM concentration (720 mg C L-1). We interpret the rapid exponential decrease in TNT as mainly an effect of sorption kinetics of TNT* and the slower linear decrease extending up to 7 d as mainly an effect of TNT degradation. In order to avoid extensive degradation of TNT and yet attain chemical equilibrium between sorbed and dissolved TNT*, we chose an equilibration time of 20 to 22 h in our equilibrium experiments.



View larger version (26K):
[in this window]
[in a new window]
  		Fig. 3. Concentrations of trinitrotoluene (TNT; filled symbols) and aminodinitrotoluene (ADNT; open symbols) in solution and 14C bound to dissolved organic matter (DOM) (filled symbols with dot) as a function of time at four different soil organic matter (SOM) concentrations. The pH was 5 to 6 and the ionic strength was held constant at 50 mM NaCl. The initial concentrations were calculated.

 
As can be seen in Fig. 2 and 3 the concentration of free ADNT increased with increasing total concentration of POM + DOM. Also the concentration of sorbed TNT* (µmol g-1 C) increased with increased concentration of POM + DOM. We suggest that an increased concentration of SOM, and its associated microorganisms, results in an increased rate of formation of reactive TNT derivatives (degradation products) and that these compounds associate more strongly with organic substances than does TNT. Recent findings using 15N-NMR show that hydroxylaminodinitrotoluene (HADNT), the precursor to ADNT in the decomposition pathway of TNT, and azoxy compounds largely are responsible for the strong bonding of TNT* to organic substances (Daun et al., 1998; Achtnich et al., 1999a). These substances were shown to form within hours under reducing conditions.

Equilibrium Experiments
Binding of TNT* to Dissolved Organic Matter in One- and Two-Phase Systems
After 20 h of equilibration, the capacity of DOM to bind TNT* per organic C atom was greater in the two-phase system than in the one-phase system (Fig. 4) . This finding is consistent with the observed increase in bound TNT* with increasing DOM + POM concentrations in the kinetic experiment (Fig. 3). Again, this may be attributed to an enhanced degradation of TNT, reflected by the increase of the free concentration of the first aminoderivative ADNT in solution. After 20 h ADNT showed a maximum of 2.6% of free TNT in the two-phase system. Still, the species bound to DOM and POM are unknown and denoted TNT*.



View larger version (23K):
[in this window]
[in a new window]
  		Fig. 4. Association of the sum of trinitrotoluene and its degradation products (TNT*) to dissolved organic matter (DOM) as a function of pH and solution concentrations of TNT. The TNT* bound to DOM was determined as 14C activity in the DOM fraction after 20 h of equilibration. Langmuir isotherms are shown for two-phase (particulate organic matter [POM] + DOM) and one-phase (DOM) systems.

 
Since the microbial activity was not monitored, it is not possible to separate biotic and abiotic degradation. Filtration of the DOM solution, however, probably reduced the number of microbes substantially, whereas extracellular enzymes can pass the filter. Thus, degradation of TNT in the one-phase system is due to activity of organisms smaller than 0.45 µm and to enzymatic and abiotic processes. Pennington and Patrick (1990) reported that the biotic reduction of TNT exceeded the abiotic reduction in soil and this is probably true in our study as well.

The isotherms for the binding of TNT* to DOM were truly nonlinear and could be satisfactorily modeled by either Langmuir or Freundlich isotherms (Table 2). The Langmuir equation gave the best fit both in one-phase and two-phase systems (0.915 < r2 < 0.980). Thus, DOM reached a saturation with respect to TNT*, which might suggest that TNT* is bound via a reaction involving a limited number of specific sites (Weber and Miller, 1989). As opposed to hydrophobic partitioning, this type of bonding is often referred to as specific and is assumed to have an ionic or covalent character. The observation that bound TNT* increased with increasing pH, disregarding small differences between pH 4.4 and 5.2 and between pH 5.6 and 6.2 in the two-phase system, might indicate that negatively charged weak acid groups in DOM are involved in the bonding.


View this table:
[in this window]
[in a new window]
  		Table 2. The sum of trinitrotoluene and its degradation products (TNT*) isotherm parameters for particulate organic matter (POM) and dissolved organic matter (DOM) at different pH. Data are both from the soil organic matter (SOM) one- and two-phase systems.

 
Carboxyl and phenol hydroxyl groups are the most numerous weak acid groups (Stevenson, 1994) in DOM (as well as in POM). Our observations are consistent with recent findings from a long-term composting experiment of an organic soil added TNT. Using 15N-NMR to characterize the bonding of TNT and its degradation products (TNT* using our designation) in fulvic acid (FA) and humic acid (HA) extracts, Achtnich and co-workers found that the association to HA had predominantly an ionic character after 4 d of reaction (Achtnich et al., 1999a). The capacity of FA + HA to bind TNT* was limited and bound TNT* (using our designation) mainly was represented by azoxy compounds. Achtnich and co-workers also suggested that the rapidly formed derivative hydroxylaminodinitrotoluene (HADNT) was condensed to azoxy species that in turn were associated with HA through ionic attraction. Since the azoxy nitrogen is electron deficient it may attract negatively charged carboxyl groups.

Binding of TNT* to Dissolved Organic Matter and Particulate Organic Matter in the Two-Phase System
Depending on experimental conditions such as pH and total concentration of TNT, 60 to 90% of the added TNT was free, 0.5 to 6% was associated with DOM, and 10 to 30% was associated with POM after 20 h of equilibration. Unidentified 14C in all experiments made up less than 3%.

The pH dependence for the sorption of TNT* to POM (Fig. 5) was much less pronounced than for the sorption to DOM. Since for DOM it was shown that ADNT formation (TNT degradation) and TNT sorption increased substantially with pH, this may indicate that TNT degradation has less importance for the short-term sorption of TNT to POM. Thus, TNT* sorption to POM probably involves quite different bonding mechanisms than to DOM.



View larger version (21K):
[in this window]
[in a new window]
  		Fig. 5. Adsorption of the sum of trinitrotoluene and its degradation products (TNT*) to particulate organic matter (POM) at different pH. The ionic medium was 50 mM NaCl. IS = ionic strength.

 
In Fig. 6 the binding of TNT* to DOM and POM is illustrated for the two-phase system at pH 5.6. For POM, Freundlich and Langmuir equations gave similar fits (0.993 < r2 < 1.000) independent on pH (Table 2). The linear Eq. [1] gave less good fits (0.932 < r2 < 0.984), especially at higher pH. Based on the Langmuir equation, the binding capacity was 15, 22, 7.3, and 6.4 times higher for POM than for DOM at pH 4.4, 5.2, 5.6, and 6.2, respectively. The isotherm for TNT* sorption to POM was more linear in shape within the experimental concentration range than the isotherm for sorption to DOM (Fig. 6). This can be evaluated by a comparison of the N values of the Freundlich equation (Table 2), which is 1.0 for absolute linearity. For POM, N varied between 0.67 and 0.80, and for DOM N varied between 0.26 and 0.34. At low concentrations of TNT the slope of the DOM isotherm was steeper indicating stronger bonding than to POM. The less pH-dependent and weaker bonding of TNT to POM, as compared with DOM, and the more linear shape and high sorption capacity of the adsorption isotherm suggest that a nonspecific form of bonding is involved in the sorption of TNT* to POM.



View larger version (20K):
[in this window]
[in a new window]
  		Fig. 6. Adsorption isotherms for the sum of trinitrotoluene and its degradation products (TNT*) associated to particulate organic matter (POM) and dissolved organic matter (DOM) at pH = 5.6 and an ionic strength of 50 mM NaCl.

 
Hassett et al. (1983) reported a linear relationship between the logarithmic forms of the experimentally determined n-octanol water partitioning coefficient (log Kow) and the organic carbon partitioning coefficient in soils (log Koc) for nonpolar organics. The equation log Koc = 0.088 + 0.909 log Kow gave an r2 of 0.92 for 34 organic compounds with a range of log Kow between 1.5 and 6.5. The relationship gives support for the idea that nonpolar compounds (having high log Kow values) are distributed between water and soil organic matter through a similar hydrophobic mechanism as in a mixture of n-octanol and distilled water. It is possible that this also holds for TNT, even though its log Kow around 1.9 indicates that TNT only partly has nonpolar properties. The apparent difference in the binding of TNT* to DOM and POM, and the fact that POM in general has a higher density of domains with hydrophobic character than DOM (Engebretson and von Wandruszka, 1997, 1998), suggests that the association of TNT* to POM may include a large contribution from hydrophobic partitioning.

At low concentration of free TNT, a relatively strong binding with ionic character at specific sites might dominate in both POM and DOM. When these sites are saturated hydrophobic partitioning take over and becomes the predominant sorption mechanism in POM. If the binding of TNT* to POM is a sum of linear partitioning to hydrophobic domains and a specific type of bonding involving polar functionality, the following simple model may be used to separate these two mechanisms:

[4]

With Eq. [4], we calculate the organic partitioning coefficient Koc at each data pair of equilibrium concentrations of TNT (Cw) and TNT* bound to POM, illustrated in Fig. 5. In the calculation we use the Langmuir values of KL and qmax determined for DOM in the two-phase system. Thus it is assumed that the specific bonding of TNT* is similar in DOM and POM at the same pH and ionic strength. The calculated Koc values are presented in Table 3. Except for the lowest concentrations of TNT the calculated Koc remained fairly constant, especially at pH 4.4 and 5.2. This indicates that an average Koc of 205 (±22) L kg-1 C may be taken as an estimate (±SD) of the constant for hydrophobic partitioning in the particulate organic matter of the soil in this study. The greater variability around the average Koc at pH 5.6 and 6.2 may be coincidental or indicates that the model (Eq. [4]) is too simple to describe the binding at higher pH.


View this table:
[in this window]
[in a new window]
  		Table 3. Sum of trinitrotoluene and its degradation products (TNT*) partition coefficients (Koc, L kg-1) for particulate organic matter (POM) calculated by Eq. [4].

 
The structure of soil organic matter changes with pH. Lower pH results in a low negative surface charge density due to protonation of COO- and ArO- groups, giving a more condensed and closed structure (Stevenson, 1994). This promotes hydrophobic partitioning reactions. The observation that the adsorption isotherm for POM becomes more linear and that the adsorption capacity shows a tendency to increase with decreasing pH (Table 2) therefore gives additional indications that hydrophobic partitioning is the main reaction by which TNT* associates with POM. The ionic strength and type of ions also influence the state of the DOM, as higher concentration and higher valence of ions leads to more closed structure with hydrophobic domains (Engebretson and von Wandruszka, 1998). Also the origin of the DOM, soil versus fluvial, affects its chemical characteristics (Myneni et al., 1999).

Values of linear organic carbon normalized partitioning coefficients (Koc) reported in the literature are listed in Table 4. Only soils with an organic content exceeding 1% are included. All values are expressed as L kg-1 C and recalculated based on data on organic C content. Our Koc values are within the range for Koc values of a variety of soils, but substantially higher than for activated carbon. Our Koc values for DOM were much smaller than the value reported for the Aldrich humic acid, the only previous study that has focused on the association of TNT to dissolved organic substances (Li et al., 1997). Unfortunately, we cannot offer any good explanation for this large difference between these two studies.


View this table:
[in this window]
[in a new window]
  		Table 4. Organic carbon partitioning coefficients (Koc) for trinitrotoluene (TNT) adsorbed to organic matter in soils and humic substances.

 
For substances without 14C labeling it would be appealing to study the binding of the substances to DOM and POM, separately. However, results from this study show that this is not possible with degradable compounds like TNT. It is obvious that the total concentration of SOM (DOM + POM) affects the degradation and subsequently the sorption processes.


   CONCLUSIONS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
REFERENCES
 
Using RP-HPLC in combination with 14C-labeled TNT, free TNT in the aqueous phase, and TNT* bound to dissolved organic matter can be satisfactorily separated and independently determined. The degradation of TNT depends on the concentration of natural soil organic substances pertaining to both the aqueous and solid phases. The chemical binding of TNT* to dissolved (DOM) and particulate soil organic matter (POM) differs significantly after 20 h of equilibration. Binding to DOM increases with TNT degradation and pH, whereas the sorption to POM is almost independent of pH. The capacity to bind TNT is much greater for POM, but the binding strength is greater for DOM, at low saturation. In general, the shapes of adsorption isotherms indicate that specific sites are involved in the bonding of TNT* in DOM whereas a more linear (hydrophobic) partitioning dominates the sorption of TNT to POM.


   ACKNOWLEDGMENTS
 
This project is financially supported by the Swedish Armed Forces. We are grateful for the help and support supplied by the personnel at the Defence Research Establishment (FOA), especially Mats Ahlberg, Jan Sjöström, Åsa Fällman, and Lars Hägglund. We also want to thank three anonymous reviewers for constructive criticism of the submitted manuscript.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 




This article has been cited by other articles:


Home page
 Soil Sci.Home page
E. G. Gregorich, M. H. Beare, U. F. McKim, and J. O. Skjemstad
Chemical and Biological Characteristics of Physically Uncomplexed Organic Matter
Soil Sci. Soc. Am. J., April 19, 2006; 70(3): 975 - 985.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (6)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Eriksson, J.
Right arrow Articles by Skyllberg, U.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Eriksson, J.
Right arrow Articles by Skyllberg, U.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by Eriksson, J.
Right arrow Articles by Skyllberg, U.
Related Collections
Right arrow Organic Compounds
Right arrow Soil Pollution
Right arrow Soil Chemistry
Right arrow Forest Soils

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
The SCI Journals Agronomy Journal Crop Science
Journal of Natural Resources
and Life Sciences Education		 Vadose Zone Journal
Soil Science Society of America Journal Journal of Plant Registrations The Plant Genome