| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Risø National Laboratory, Plant Research Department, Building PRD-301, Post Office Box 49, DK-4000 Roskilde, Denmark
b The Royal Veterinary and Agricultural University, Chemistry Department, Thorvaldsensvej 40, DK-1870 Frederiksberg, Denmark
c The Danish University of Pharmaceutical Sciences, Department of Analytical Chemistry, Universitetsparken 2, DK-2100 Copenhagen, Denmark
* Corresponding author (amja{at}dfh.dk).
Received for publication July 28, 2002.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Abbreviations: HPLC, high performance liquid chromatography • LAS, linear alkylbenzene sulfonate • NP, nonylphenol • NPE, nonylphenol ethoxylate • SPE, solid-phase extraction
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
Surfactants can affect soil microorganisms and invertebrates by dissolving biomembranes (Jensen, 1999). Toxicological effects on several microbial parameters have been demonstrated and EC50 (effective concentration for 50% of the test population) values ranging from 1143 to 1437 mg kg–1 dry wt. for LAS and 71 to 437 mg kg–1 dry wt. for NP have been reported for the soil invertebrates Folsomia candida and Enchytraeus albidus (Gejlsbjerg et al., 2001). However, it was estimated that LAS does not pose a significant risk to soil fauna or plants, given normal sludge amendment (Jensen et al., 2001). In contrast, it has been shown that NP has estrogenic effects. Studies using rainbow trout (Oncorhynchus mykiss) have revealed a reduction in testicular growth at concentrations of 1 µg L–1, while complete growth inhibition was observed at 54 µg L–1 (Jobling et al., 1996).
The surfactants can also affect the environment by mobilizing organic hydrophobic compounds in soil. At concentrations above the critical micelle concentration (Table 1), the surfactants form micelles into which organic compounds partition. These micelles are mobile and transported with soil water (Danzer and Grathwohl, 1998).
Several studies have shown relatively rapid degradation of LAS and NP. Reported half-lives for LAS range from 3 d in a lysimeter study using a sandy soil (Küchler and Schnaak, 1997) to 7 to 22 d in a study of a large variety of field soils (Holt et al., 1989). Half-lives for NP have been determined to be 20 d in a laboratory study (Staples et al., 1999) and 4.5 to 16.7 d in a study with six different soils (Topp and Starratt, 2000).
The sorption of LAS to soil is a combination of several mechanisms and sorption to both the organic and inorganic fraction of the soil has been demonstrated (Küchler and Kujawa, 1998). The linear alkyl group of LAS is hydrophobic and sorbs to the nonpolar fractions of the soil, such as the organic matter (Küchler and Kujawa, 1998). However, the sulfonate group of LAS is negatively charged and hydrophilic and therefore interacts with positively charged soil components or polar groups, such as hydroxy-groups, minerals, or oxides (Ou et al., 1996). However, the negatively charged sulfonate group is repelled by negatively charged surfaces, and hence the sorption of LAS has been demonstrated to decrease with increasing pH (Fytianos et al., 1998; Westall et al., 1999). There is no clear correlation between the type and content of soil components and the extent of LAS sorption. However, some studies show a correlation between clay content and sorption (Ou et al., 1996; McAvoy et al., 1994), while other studies show correlation between the content of organic matter and sorption (Fytianos et al., 1998; McAvoy et al., 1994; Painter, 1992).
The sorption mechanism for the nonionic surfactant NP is much simpler. The compound is only deprotonated when pH exceeds the pKa value for phenols of approximately 10. Consequently, under most environmental conditions the molecule has no electrical charge and therefore primarily interacts with the soil by hydrophobic sorption to the organic fraction. However, sorption by hydrogen bonding also has been demonstrated (John et al., 2000).
Partitioning coefficients (Kd) for sorption to soil, sludge, and other relevant matrices of LAS and NP are listed in Table 2. Values for NPE are also included, since only few data are available for NP. The values for NPE and NP are relatively high, indicating strong sorption to the solid phase of the soil. However, the values for LAS range around 1 L kg–1, indicating an equal distribution of LAS between the soil water and the solid phase, depending on the water content. Thus, it is expected that a large fraction of LAS applied to soil with sewage sludge is present in the soil water and the potential for leaching of LAS is assumed to be relatively high, whereas the mobility of NPE is much lower because of strong sorption to soil particles. However, for both surfactants sorption to sludge and other matrices with high content of organic C is very strong, which may retain the compounds in sludge-amended soil.
|
The aim of this study was to estimate, by use of lysimeters, the degradation and mobility of LAS and NP in soil after application of sewage sludge to agricultural soils. The experimental lysimeter design approached realistic field conditions, because the surfactants were applied in realistic concentrations with sewage sludge in realistic amounts immediately before the growth season. The lysimeters were grown with crops (spring barley) and the amount of water applied to the lysimeters was not unrealistically high.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
|
|
On all lysimeters 30 barley seeds were sowed and fertilizer (NPK 21–3–10; KEMIRA, Fredericia, Denmark) was applied in a dosage equivalent to 117 kg N ha–1 as a supplement to the N and P added with the sewage sludge (Table 4).
During the experimental period of 110 d the lysimeters received 300 mm of natural precipitation, but were also artificially watered with tap water to force water through the column. By the end of the experiment the total amount of water applied to each lysimeter was 53 L or 700 mm of precipitation, which is the average yearly rainfall in Denmark.
The soil temperature was not measured during the experimental period, but because the lysimeters were placed underground, it is expected that the temperature was well below air temperature, but probably higher than under typical field conditions.
Sampling Strategy
Soil samples were collected from three depths (0–15, 15–30, and 30–45 cm). Due to the relative small size of the lysimeters, it was not possible to take soil samples by soil auger or otherwise take samples from all soil layers without disturbing the complete soil column and hence change the water flow in the columns. Therefore, three lysimeters were taken out of the experiment at each sampling day. Each complete soil column was dissected into the three layers and each layer was mixed to homogeneity before sampling. Hence, 15 lysimeters gave the possibility for five sampling days with three replicates, which were determined to be Day 10, 20, 30, 50, and 110, with more frequent sampling in the beginning of the experimental period, due to short half-lives demonstrated in previous works. Therefore, each sampling day represents soil samples from three replicate lysimeters with samples from three soil layers.
Leachate samples were collected in polyethylene bottles on Days 0, 10, 20, 30, 50, 70, and 110 by emptying the leachate collection bottles using an electrical pump. Since three lysimeters were taken out of the experiment at each day of soil sampling, the number of replicates of leachate samples declined during the experimental period from 15 replicates on Days 0 and 10 to three replicates on Days 70 and 110. Furthermore, at each sampling day leachate samples were collected from the three control lysimeters without sludge application. The volume of leachate collected varied between lysimeters and from day to day. On average a total of 5 L or 66 mm were collected per lysimeter during the 110 d, which is approximately 10% of the water applied. Most leachate was collected in the first and last period of the experiment, correlating with the low biomass of the young and ripe barley plants, respectively. This indicates high evapotranspiration from the lysimeters during the growth period, since the collection bottles were completely sealed except for the 1.5-m-long Teflon tube (see Fig. 1), through which evaporation is assumed to be negligible.
Leachate and soil samples were stored at –18°C until analysis.
Leachate Analysis
Linear alkylbenzene sulfonate was extracted without pretreatment from 50 mL of the collected leachate samples on C8 (500 mg of sorbent and a 6-mL cartridge) SPE columns (Supelco, Bellafonte, PA). The SPE columns were conditioned with 15 mL methanol followed by 15 mL deionized water. After sample extraction LAS was eluted with 3 x 3 mL methanol and evaporated at 50°C to 1 mL under a gentle stream of nitrogen. Quantification of LAS was achieved using a Shimadzu (Kyoto, Japan) HPLC system with fluorescence detection. The HPLC column was a Supelcosil LC-8-DB column with dimensions of 25 cm x 4.6 mm and particle size of 5 µm. The mobile phase was a two-solvent gradient, consisting of HPLC-grade water and acetonitrile. From the time of injection to 10 min the volumetric composition was 30% acetonitrile and 70% water, then followed a linear gradient to 99% acetonitrile at 50 min. Flow was 1 mL min–1 and the temperature was 25°C. The HPLC method separated the homologs C10, C11, C12, and C13 LAS.
Nonylphenol was likewise extracted from the leachate samples by SPE using ENV+ SPE columns (50 mg of sorbent, 10-mL cartridge) from Isolute (International Sorbent Technology, Mid Glamorgan, UK). Before extraction each 50-mL leachate sample was adjusted to pH 2 to 3 using sulfuric acid and 250 µL methanol was added. The SPE columns were conditioned successively with 6 mL acetone, 10 mL methanol, and 6 mL deionized water (pH = 2). Nonylphenol was eluted with 2 x 3 mL acetone from the ENV+ columns and evaporated at 20°C and under a gentle N2 stream to a volume of 300 µL, followed by dilution with propanol to a volume of 2 mL. Quantification of NP was performed using HPLC with fluorescence detection and isocratic elution (30% water and 70% acetonitrile) with a flow of 1 mL min–1 at 25°C. The HPLC column was an Envirosep-pp (12.5 cm x 4.60 mm) from Phenomenex (Torrance, CA). The method separated NP from NPE, but did not separate different isomers of NP and therefore quantified all branched and linear NP isomers in one signal.
Linear alkylbenzene sulfonate and NP were analyzed in series of 10 leachate samples plus one recovery sample and one repeatability sample. To determine recovery, one sample per series was fortified with a known amount of surfactant, that is, a solution of the technical LAS product Isorchem (Condea, Paderno Dugnano, Italy) with known content of LAS homologs and unbranched NP (n-NP), respectively. Recovery for the method was determined as the average value (and corresponding standard deviation) for recovery determined in all series. For repeatability one sample in each series was analyzed in duplicate, and repeatability for the method was determined as the average deviation between duplicate measurements. Furthermore, one blind sample with HPLC-grade water was analyzed for each series, resulting in a total of 13 samples per series.
An overview of the analytical methods used for leachate samples and the method validation parameters is presented in Table 5.
|
The extraction of NP from soil was performed without pretreatment of the soil samples. Samples of approximately 20 g soil or less than 2.5 g sludge were each mixed with 30 mL water and 25 mL cyclohexane and acetone (90:10). The samples were shaken mechanically for 16 h, centrifuged, and extracted on NH2 (aminopropyl) (500 mg of sorbent, 6-mL cartridge) SPE columns from Supelco, conditioned with 15 mL cyclohexane. After sample extraction NP was eluted with 2 x 3 mL cyclohexane and acetone (60:40), evaporated at room temperature under a N2 stream to a volume of 300 µL, and diluted to 2 mL with propanol. Quantification of NP was performed by HPLC, as for leachate samples.
Linear alkylbenzene sulfonate as well as NP were analyzed in a series of nine soil samples representing one sampling day (three soil layers from three lysimeters). In each series recovery was determined by fortifying one soil sample with a known amount of surfactant before extraction. Obviously, fortification shortly before extraction may overestimate the extraction efficiency, but contrary degradation of surfactants may take place if left in the soil mixture for a longer period. Repeatability was determined by performing duplicate extractions on a sample from the 0- to 15-cm soil layer. Furthermore, one blind sample without soil was analyzed for each series, resulting in a total of 12 samples per series. The chemicals and water used for soil extractions, SPE and HPLC analysis were all HPLC grade.
|
RESULTS AND DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
The measured average contents of LAS and NP in the top-15-cm soil layer throughout the 110-d experimental period are illustrated in Fig. 2
. Each point with corresponding error bars represents the average concentration and standard deviation of four measurements, that is, the results from three lysimeters plus one duplicate measurement. The fitted curves are logarithmic trend lines, corresponding to the equations:
 |
 |
|
In the soil layers below the 15 cm of sludge incorporation, none of the surfactants were measured in concentrations above the detection limits of 150 and 50 µg kg–1 dry wt. for LAS and NP, respectively. These results suggest that none of the surfactants were transported out of the top-15-cm soil layer of sludge incorporation during the experimental period (see below for discussion of mobility).
Degradation and Half-Lives
A rapid decline in the soil contents of both LAS and NP in the top-15-cm soil layer was observed within the first 10 d. This was followed by a continuous but slower decline (Fig. 2). Several authors have demonstrated first-order degradation kinetics for LAS (Krueger et al., 1998; Larson et al., 1989) and NP (Staples et al., 1999; Kvestak and Ahel, 1995). However, for this study, first-order degradation kinetics for the complete experimental period resulted in poorly fitted degradation curves for both surfactants (r2 = approximately 0.8), particularly due to the relatively fast initial degradation in the period from the start of the experiment to the first sampling day (Day 10). According to Madsen et al. (1999), the kinetics of degradation of organic compounds in sludge-amended soils (phthalates) can be divided into two phases with different kinetic expressions representing initial and long-term degradation, respectively. Hence, the results of this study were divided into a first phase represented by Days 0 to 10. In this period the initial soil content of LAS was reduced by 75% (to 9000 µg kg–1 dry wt.), while the content of NP was reduced by 55% (to 256 µg kg–1 dry wt.), indicating initial half-lives of less than 10 d. However, too little data is available for kinetic analysis of this period, and hence this period was left out of estimations of half-lives.
First-order degradation kinetics were assumed for this study and data from Days 10 to 110 were used to estimate half-lives for the surfactants. Degradation (disappearance) in this period was much slower and by the end of the experimental period (Day 110) the initial soil content of LAS was reduced by more than 99% (to 224 µg kg–1 dry wt.) and the NP content was reduced to below the detection limit (<50 µg kg–1 dry wt.).
Inserted in Fig. 2 is a first-order degradation plot, where the surfactant concentration is plotted logarithmically as a function of time. Regression analysis on these curves yields degradation rate constants, k, of 0.01512 and 0.00812 d–1 for LAS and NP, respectively, and corresponding half-lives, t1/2, of 20 (16–25) days for LAS and 37 (31–46) days for NP (95% confidence levels shown in parentheses).
The degradation patterns for both surfactants found in the top-15-cm soil are similar to previous studies, with a rapid initial degradation followed by a slower secondary degradation (Marcomini et al., 1989). The half-lives estimated in this study for LAS and NP are compared with previous works in Table 6. In most cases a satisfactory agreement is seen (Holt et al., 1989; Berna et al., 1989; Topp and Starratt, 2000), except for some studies where the half-lives are as short as a few days (Küchler and Schnaak, 1997). However, these very short half-lives may correspond to the fast degradation demonstrated in the first 10 d of the lysimeter experiment.
|
By the end of the experimental period of 110 d the surfactants had almost disappeared from the soil, indicating that the surfactants will not accumulate in sludge-amended agricultural soils from year to year. However, in this study no sludge aggregates were present, as the sludge was well incorporated in the soil, which stimulates fast degradation of the surfactants compared with compounds trapped in larger sludge aggregates. Therefore, degradation in this study may be overestimated compared with field conditions.
Relative Distribution of Linear Alkylbenzene Sulfonate Homologs
During the 110-d experimental period there was a change in the relative distribution of LAS homologs in the top-15-cm soil layer. The fraction of the C13 homolog increased from 35 to 64%, while the fractions of the three other homologs declined (Fig. 3)
.
|
This is supported by comparing with the commercial formulation of LAS (Isorchem), where the relative distribution of homologs is 14% C10, 34% C11, 31% C12, and 21% C13 (Fig. 3), and hence assumed to be the relative distribution in the wastewater reaching the wastewater treatment plant. The sludge used in this study contained only 3% of the C10 homolog, indicating that the homolog shift takes place already at the wastewater treatment plant and it is primarily the C13 homolog that is accumulated in the sewage sludge (Table 4).
Mobility in Soil
Neither of the surfactants were measured in concentrations above the analytical detection limit in any of the water samples collected from the lysimeter leachate. Furthermore, neither LAS nor NP were detected in concentrations above the detection limit in soil layers below the depths of sludge incorporation (15 cm). These results indicate negligible downward transport of the surfactants in the lysimeters.
The lack of downward transport cannot be explained solely by the rapid degradation of the surfactants immediately after application of sludge to the soil. Although the contents of surfactants are lowered to near the detection limit by the end of the study, the soil water concentration for most of the study is expected to be well within the detectable level, particularly for LAS. Since the first sampling day is after 10 d, there is a possibility that surfactants leached to the collection bottles during the first days are degraded before the first sampling, as bioavailability is higher in the water samples than in the soil. However, such downward transport is expected to result in detectable soil concentrations of the surfactants in the soil layers below 15 cm, unless degradation is fast in these soil layers and half-lives are considerably shorter than estimated above for the sludge-mixed soil layer (0–15 cm).
In addition, the flow of water through the lysimeters during the experimental period was limited. Only 10% of the water applied (precipitation and irrigation) passed through the 45-cm soil column and was collected as leachate. Most leachate was collected in the first and last period of the experiment, correlating with the lower evapotranspiration from the young and ripe barley plants, respectively. This situation corresponds well with field conditions, where the water balance during the growth season generally is negative for cropped fields due to high evapotranspiration, resulting in upward water transport in the soil. The annual potential evapotranspiration from Danish farmland grown with cut grass (reference crop) is approximately 525 mm (Jensen et al., 1997), and was possibly higher from the lysimeters due to slightly higher soil temperature and more frequent precipitation and irrigation. Hence, although only little water transport is seen through the lysimeters, leaching of the surfactants has been stimulated in this study compared with field conditions.
The major reason for lacking mobility of NP is expected to be strong sorption to soil and sludge, while sorption coefficients of approximately 1 L kg–1 for sorption of LAS to soil (Table 2) indicate that sorption to soil particles is not strong and it is therefore surprising that no transport of LAS out of the top-15-cm soil layer is seen. However, the surfactants were added to the soil with sewage sludge and Kd values of up to 12000 L kg–1 for sorption to sludge (Table 2) indicate very strong sorption compared to soil. It has been demonstrated that the mobility of NP from sludge aggregates to the surrounding soil is negligible (Hesselsøe et al., 2001) and even though the sludge is well incorporated in the soil in this study, the surfactants may still be strongly sorbed to the sludge particles, retaining the compounds in the top soil layer. In contrast, sludge colloids may be mobile in soil and colloid-facilitated transport of the surfactants is possible, simply with soil water or through macropores present in agricultural fields. In this study colloid-facilitated transport was limited due to low downward flow of soil water, and furthermore the soil columns contained packed soil rather than intact soil cores, reducing preferential flow (e.g., through macropores).
The aim of this lysimeter study was to mimic field conditions and the results of this study, showing no downward transport of the surfactants, differ from those obtained in other experiments. Transport of LAS through soil columns has been observed in two other lysimeter studies using 40- or 10-cm soil columns with sandy soils, and leaching has been demonstrated in soils with or without organic matter (Küchler and Schnaak, 1997; Küchler and Kujawa, 1998). Furthermore, the work of Ou et al. (1999) demonstrated leaching of LAS through 1.5-m soil columns with clayey soil. In addition, several authors have observed downward transport of NP. Ahel et al. (1996) and Barber et al. (1988) have described the downward transport of NP in soils below a highly contaminated river and a wastewater pond, respectively, while Vikelsøe et al. (1999) have measured NP at depths of 60 cm below a sludge-amended field (17 Mg dry wt. per hectare per year). However, in these experiments, the surfactants are generally added to the soils in high concentrations (for instance the 400 mg LAS kg–1 dry wt. in the experiment by Ou et al., 1999) and often dissolve in either water or methanol, whereas the present lysimeter study represents more realistic conditions. The surfactants were added in realistic concentrations with sewage sludge, to which sorption is strong and therefore mobility reduced. Furthermore, realistic amounts of water (precipitation) were applied to the lysimeters and barley was grown throughout the experimental period, resulting in high evapotranspiration and therefore limited downward transport of soil water. The results of this lysimeter study indicate negligible downward transport of surfactants added with sewage sludge immediately before the growth period, where upward water flow dominates and where elevated temperatures stimulate fast degradation. However, under field conditions degradation may be slower due to low oxygen content in larger sludge aggregates and leaching may happen as colloid-facilitated transport through macropores. Furthermore, it is possible that the surfactants are mobile in soil during autumn and winter, with low temperatures, no or little plant growth, and downward water transport.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
In the top soil layer an initial rapid degradation of both surfactants was observed, followed by slower but continuous degradation. After the first 10 d of the study, the contents of the surfactants were only 25 and 45% of the initial content for LAS and NP, respectively. By the end of the experiment, after 110 d, less than 1% LAS was left and the content of NP was below the analytical detection limit.
Assuming first-order degradation kinetics, half-lives of 20 and 37 d for LAS and NP, respectively, were estimated. The half-lives for the C13 LAS homolog were higher than for the lower homologs, resulting in a relative accumulation of this homolog in the soil during the experiment.
The lysimeter study showed no downward transport of LAS or NP, as the surfactants were not measured in the leachate samples or in soil layers below the top-15-cm soil layer with sludge incorporation at any time during the 110-d experimental period.
|
ACKNOWLEDGMENTS |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 | |||
