Environ. Sci. Technol. 2003, 37, 624-629
Induced Phytoextraction/Soil Washing of Lead Using Biodegradable Chelate and Permeable Barriers B O Sˇ T J A N K O S A N D D O M E N L E Sˇ T A N * Agronomy Department, Biotechnical Faculty, University of Ljubljana, Jamnikarjeva 101, 1000 Ljubljana, Slovenia
Chelate-induced remediation has been proposed as an effective tool for the extraction of lead (Pb) from contaminated soils by plants. However, side-effects, mainly mobilization and leaching of Pb, raise environmental concerns. Biodegradable, synthetic organic chelate ethylenediaminedisuccinic acid (EDDS), and commonly used ethylenedimanetetraacetic acid (EDTA) were used for induced phytoextraction with a test plant Brassica rapa and in situ washing of soil contaminated with 1350 mg/kg of Pb. Horizontal permeable barriers were placed 20 cm deep in soil columns and tested for their ability to prevent leaching of Pb. The reactive materials in the barriers were nutrient enriched vermiculite, peat or agricultural hydrogel, and apatite. EDTA and EDDS addition increased Pb concentrations in the test plant by 158 and 89 times compared to the control, to 817 and 464 mg/kg, respectively. In EDTA treatments, approximately 25% or more of total initial soil Pb was leached in single cycle of chelate addition. In EDDS treatments, 20% of the initial Pb was leached from columns with no barrier, while barriers with vermiculite or hydrogel and apatite decreased leaching by more than 60 times, to 0.35%. 11.6% of total initial Pb was washed from the soil above the barrier with vermiculite and apatite, where almost all leached Pb was accumulated. Results indicate that use of biodegradable chelate EDDS and permeable barriers may lead to environmentally safe induced Pb phytoextraction and in situ washing of Pb.
Introduction Metals, including lead, chromium, arsenic, zinc, cadmium, copper, and mercury, can cause significant damage to the environment and human health. Remediation of sites contaminated with toxic metals is particularly challenging. Unlike organic compounds, metals cannot be degraded, and cleanup requires their immobilization and toxicity reduction or removal. The selection of the most appropriate soil remediation method depends on the site and soil characteristics and heavy metals concentration. Solidification/ stabilization contains the contaminants in an area by mixing or injecting agents such are cement and lime. Electrokinetic processes involve passing a low intensity electric current between a cathode and an anode imbedded in the con* Corresponding author phone: +386 61 123 1161; fax: +386 61 123 1088; e-mail:
[email protected]. Corresponding address: Center for Soil and Environmental Science, Biotechnical Faculty, University of Ljubljana, Jamnikarjeva 101, 1000 Ljubljana, Slovenia. 624
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 37, NO. 3, 2003
taminated soil. An electric gradient initiates charged particle movement. The metals can be removed by electroplating or precipitation at the electrodes. Soil washing involves the addition of water with chelates such as ethylenediaminetetraacetic acid (EDTA) or acids in combination with chelates. This can be done in reactors or as a heap leaching. Heavy metal contaminated sandy soils (less than 10-20% clay and organic matter) are most effectively remediated (1). Many of these technologies have been used full-scale but are an energy-intensive approach and can be prohibitively expensive. Stringent physical treatments and chemical agents involved can dramatically inhibit soil fertility, with subsequent negative impacts on the ecosystem (2). Phytoextraction of metals uses metal-accumulating plants to clean up contaminated soils. Phytoextraction is an aesthetically appealing green technology that preserves or even enhances soil quality. Harvestable parts of the plants, rich in accumulated metals, can be easily and safely processed by drying, ashing, or composting. Some extracted metals can also be reclaimed from the ash, generating recycling revenues (3). Plants such as Thlaspi, Urtica, Chenopodium, Polygonum sachalase, and Alyssim have the capability to hyperaccumulate Cd, Cu, Pb, Ni, and Zn (1, 4), but their biomass production is generally considered too low to make these plants feasible candidates for phytoextraction. Chelate-enhanced phytoremediation has been proposed as an effective tool for the extraction of heavy metals from soils by plants. Grcˇman et al. (5), Barona et al. (6), and many others have reported that after EDTA soil treatment, the proportion of Pb accumulated in the water and acidextractable fraction considerably increased, which was related to the greater degree of metal extraction from the other soil fractions. As a consequence, after EDTA application, the metal content remained more weakly adsorbed to soil components (more easily leachable), potentially favoring the application of phytoremediation technologies. Chelate mobilization is especially important for phytoextraction of Pb. Of all toxic heavy metals commonly occurring in contaminated soil, Pb is the least phytoavailable. Lead in soil is strongly complexed with organic matter, sorbed on oxides and clays, and precipitated as carbonate, hydroxide, and phosphate (7). In every contaminated soil we have examined with sequential extractions, none or a very small fraction of total soil Pb was present in the soil solution (8). The literature to date reports a number of chelates that have been tested for chelate-induced hyperaccumulation of heavy metals, particularly Pb. These include EDTA and its structural analogues: trans-1,2-diaminocyclohexane-N,N, N′,N′-tetraacetic acid (CDTA), dietylenetriaminepentaacetic acid (DTPA), etyleneglycol-bis(β-aminoethyl ether) (EGTA), etyleneglycol-bis(β-aminoethyl ether) (EDDHA), N-hydroxyethylenediaminetriacetic acid (HEDTA), nitrilotriacetic acid (NTA) (9-11). In most cases, the EDTA treatment was superior in terms of solubilizing soil Pb for root uptake and translocation into above-ground biomass. In a maize crop, EDTA 0.2% treatment resulted in 2435 mg/kg Pb in shoot tissues (12). Blaylock et al. (9) tested five chelating agents. They used 3 week old seedlings and measured more than 15 000 mg/kg Pb (300-fold increase) in the dry weight of shoots of Brassica juncea after 10 mmol/kg EDTA addition. Huang et al. (13) determined 8960 mg/kg (111-fold increase) and 2410 mg Pb/ kg (57-fold increase) in 2 week old pea and corn shoots transplanted into soil substrate pretreated with 1.5 mmol/ kg EDTA. However, side-effects related to the addition of chelates, e.g. metal leaching are expected and cause additional health, 10.1021/es0200793 CCC: $25.00
2003 American Chemical Society Published on Web 01/01/2003
safety, and environmental concerns. Many synthetic chelates and their complexes with heavy metals are toxic (14) and poorly photo-, chemo-, and biodegradable in soil environments (15). Romkens et al. (16) conducted lysimeter studies and reported that EDTA analogue EDGA enhanced metal leaching, which could lead to groundwater pollution. Grcˇman et al. (17) reported that in a soil column experiment 37.9, 10.4, and 56.3% of initial total Pb, Zn, and Cd were leached out of the soil after treatment with 10 mmol/kg EDTA. Further research and development of environmentally safe methods of chelate induced phytoextraction is needed before chelateassisted phytoextraction techniques can be widely used at contaminated sites. In this study, the biodegradable, low-toxic synthetic organic chelate, [S,S]-stereoisomer of ethylenediaminedisuccinic acid (EDDS), was used for induced phytoextraction. EDDS was first isolated as a metabolite of the soil actinomycete Amycolatopsis orientalis (18). It is therefore naturally present in soil, where it is easily decomposed into benign degradation products (19). The biodegradability of metalEDDS complexes were studied by Vandevivere et al. (20). Activated sludge fed with EDDS as the sole C and N source was shown to readily biodegrade 1 mM concentration of Ca-, Cr(III)-, Fe(III)-, Pb-, Al-, Cd-, Mg-, Na-, or ZnEDDS. On the other hand, the Cu-, Ni-, Co-, and HgEDDS complexes remained essentially undegraded. Only in the case of Hg-EDDS was the lack of biodegradation due to metal toxicity. Jaworska et al. (21) assessed environmental risks for use of EDDS in detergent applications. Mineralization of EDDS in sludge-amended soil was rapid and complete in 28 days. The reported calculated half-life was 2.5 days. The toxicity to fish and daphnia was low (EC50 > 1000 mg/L). EDDS has been substituted for traditional chelates in a number of commercial products, e.g., industrial detergents. The current price for 1 ton of EDDS is approximately 5000 GBP. Vandevivere et al. (22) investigated EDDS for its applicability for the ex situ washing extraction of Pb, Zn, Cu, and Cd from soil, sewage sludge, and harbor sediments. They reported that it was feasible to achieve 70-90% extraction of Zn, Pb, and Cu from the solids tested. Extraction efficiencies were equal or superior to those obtained with the benchmark chelate EDTA. There are no reports on using EDDS or other chelates for in situ soil washing. For reducing mobilization of toxic metals and organic pollutants in groundwater at contaminated sites vertical permeable reactive barrier are being evaluated (1). A permeable reactive barrier is constructed below ground as a vertical underground wall, filled with reactive materials. Barrier is built by digging a long, narrow trench in the path of the polluted groundwater. Sometimes the barrier is part of a funnel that direct the polluted groundwater to the reactive part of the wall. Clean groundwater flows out of the other side of the wall. Various reactive materials have been investigated and include zeolite, elemental iron, and limestone. Reactive materials in the barrier trap harmful chemicals or change the chemicals into harmless ones. For example, zerovalent iron can be used for reduction of toxic Cr6+ to harmless Cr3+ and limestone for Pb participation (23). We borrowed this concept to propose a horizontal permeable barrier to prevent leaching of the heavy metal-chelate complex through the soil profile in methods of chelate induced phytoextraction and in situ soil washing of heavy metal contaminated soil. The objective of our study was to evaluate enhanced phytoextraction of Pb supported by addition of commonly used chelate EDTA and biodegradable chelate EDDS, combined with in situ soil washing of Pb using the same chelates and permeable barriers to minimize losses of Pb. In a laboratory study, horizontal permeable barriers were placed 20 cm deep in soil columns and tested for their ability to
TABLE 1. Selected Physical and Chemical Properties of the Soils Used in This Study soil property pH (CaCl2) COX (%) total N (%) P as P2O5 (mg/100 g) K as K2O (mg/100 g) sand (%) coarse silt (%) fine silt (%) clay (%) total Pb (mg/kg)
6.8 5.2 0.25 37.3 9.2 55.4 12.0 18.9 13.7 1350
Pb fractionation (%) in soil solution exchangeable from colloids bound to carbonate bound to Fe and Mn oxides bound to organic matter residual fraction
0.1 0 36.6 0.4 51.1 11.8
prevent leaching of Pb, mobilized by the addition of chelates. The reactive materials used in the permeable barriers were nutrient enriched vermiculite, peat, or agricultural hydrogel, in some treatments followed by a layer of apatite/soil mixture. Leaching and distribution of Pb through the soil profile and Pb uptake by the test plant Brassica rapa were determined.
Materials and Methods Soil Properties. Soil samples were collected from the 0-30 cm surface layer at an Pb contaminated industrial site of a former Pb smelter in Slovenia. The soil texture was sandy loam. Selected physical and chemical properties of the soil are presented in Table 1. A modified sequential procedure according to Tessier et al. (24) was used to determine fractionation of Pb into six fractions (5). In the first and second fraction (soluble in soil solution and exchangeable from soil colloids) 0.1% of total Pb was determined; Pb from the first two fractions is assumed to be bioavailable to plants. The rest of Pb was determined in fractions not directly available to plants (Table 1). Experimental Set Up. After being air-dried, the soil was passed through a 4-mm sieve. EDDS (Octel, Cheshire) and EDTA (Fluka) induced Pb plant uptake, and washing and leaching were studied in a soil column experiment with four replicates for each of the following treatments: control columns with no horizontal permeable barriers (EDDS and EDTA treatment); columns with horizontal permeable barriers composed of a layer of nutrient enriched substrates selected from hydrogel, vermiculite or peat (EDDS and EDTA treatment); columns with horizontal permeable barriers composed of a double width layer of selected nutrient enriched substrate (EDDS treatment only); columns with horizontal barriers composed of a layer of selected nutrient enriched substrate followed by a layer of apatite/soil mixture (EDDS treatment only). For control, we placed 3755 g of air-dried soil into 18 cm high, 15 cm diameter columns. The columns with horizontal permeable barriers were 27 cm high and of the same diameter. Permeable barriers were positioned 20 cm below 4400 g of air-dried soil. The layer of enriched substrates was 1.5 or 3 cm wide (single and double layer) and was followed by a 3 cm wide layer of soil in the bottom of the column. In horizontal permeable barrier treatments with apatite layer (1.5 cm wide), this layer followed the 1.5 cm enriched substrate layer and 1 cm soil layer. The apatite layer was followed by a 3 cm soil layer in the bottom of the column. VOL. 37, NO. 3, 2003 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
625
All columns were equipped with trapping devices for leachate collection. Plastic mesh (D ) 0.2 mm) was placed between layers and at the bottom of the columns to retain the soil. As carrier materials for nutrient enriched substrates, 7 or 14 g (single or double layer) of acrylamide hydrogel Stocksorb K 400 (Stockhausen, Krefeld), 40 or 80 g of commercial vermiculite, and 35 or 70 g of commercial peat were used. These materials were soak-enriched with glucose (100 g), yeast extract (15 g), KH2PO4 (3 g), and (NH4)2SO4 (3 g) by placing them in jars with a solution of appropriate nutrient concentration in a volume of water sufficient for 50% water saturation of materials. The apatite layer was prepared by mixing 12 g of synthetic apatite (Riedel-de Hae¨n, Seelze) with 270 g of dry soil. Soils were fertilized in all treatments with 100 mg/kg N and K as (NH4)2SO4 and K2SO4, respectively. Three week old seedlings of Brassica rapa L. var. pekinensis (Nagaoka F1) were transplanted into the columns. EDDS and EDTA were applied in 200 mL of deionized water in a single dose of 10 mmol/kg soil on the 27th day of cultivation. The aboveground tissues were harvested on the 30th day of cultivation by cutting the stem 1 cm above the soil surface. Biomass was determined after the tissues were dried at 60 °C and reached a constant weight. Soils were watered 2 times a week with 250 mL of tap water. Leachates were sampled once a week for 9 weeks after EDDS and EDTA application. They were filtered through Whatman No. 1 filter paper and stored in cold storage for further analysis. Water Sorption Capacity of Hydrogel, Vermiculite, and Peat. We added 0.2 g of hydrogel, vermiculte, or peat into the jar with 100 mL of water. After a 24 h incubation at 25 °C, the saturated materials were removed from the jar and weighed. They were then dried at 60 °C until a constant weight was reached and weighed again. The water sorption capacity of materials was expressed as a percentage of water sorbed. Microbial Activity in Horizontal Barriers. Microbial activity in the horizontal barriers was determined as the metabolic heat generated in a separate set of four insulated, 27 cm high soil columns (one for the control, three with horizontal permeable barriers consisting of a single layer of acrylamide hydrogel, vermiculite or peat, enriched with substrates). No plants were planted into these columns. Barriers were positioned 20 cm below the soil surface and were 1.5 cm wide. Temperature probes were inserted 21 cm deep into the horizontal barriers and into the soil of the control column. Heavy Metal Determination. For the analysis of metal concentrations, soil and vermiculite samples were ground in an agate mill for 10 min and then passed through a 150 µm sieve. After digestion in aqua regia, AAS was used for the determination of metal concentrations. Shoot tissues were collected and thoroughly washed with deionized water. They were dried to a constant weight and ground in a titanium centrifugal mill. Metal concentrations in plant tissue samples (290-310 mg dry weight) were determined using an acid (65% HNO3) dissolution technique with microwave heating and analyzed by Flame-AAS. Heavy metal concentrations in leachates were determined by FlameAAS. Controls of the analytical procedure were performed using blanks and reference materials (BCR 60 and BCR 141R, Community Bureau of Reference, for plant and soil) that were treated identically to experimental samples. Two measurements of heavy metals were performed for each sample. Statistical Analyses. To allow analysis of variance, we made four replicates of each treatment. When necessary, the data were log-transformed before statistical analysis to stabilize the variance. The LSD multiple range test was used to determine the statistical significance (P ) 0.05) between 626
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 37, NO. 3, 2003
TABLE 2. Percent of Total Initial Pb Accumulated in Plant Above-Ground Parts and Leached from Soil after 10 mmol/kg Chelate Addition, during Nine Weeks of Soil Irrigation with Waterd chelate control EDDS EDTA barrier
chelate
peat peat (double layer) peat + apatite peat hydrogel hydrogel (double layer) hydrogel + apatite hydrogel vermiculite vermiculite (double layer) vermiculite + apatite vermiculite no barriereb no barriereb control
EDDS EDDS EDDS EDTA EDDS EDDS EDDS EDTA EDDS EDDS EDDS EDTA EDDS EDTA
total Pb accumulated (%) 0.001 ( 0.000 0.03 ( 0.01 0.05 ( 0.02 total Pb leached (%) 2.10 ( 1.73a 2.00 ( 0.76a 0.62 ( 0.87a 23.96 ( 5.98 1.57 ( 0.76a 1.61 ( 1.31a 0.35 ( 0.31a 25.26 ( 2.07 2.59 ( 0.59a 1.48 ( 1.76a 0.34 ( 0.40a 27.12 ( 1.53 21.52 ( 6.13 36.17 ( 1.32 NDc
a Statistically different EDDS treatments according to LSD test (P ) 0.05). b Measurement 4 weeks after chelate addition. c ND, not detected. d Results are presented as means of four replicates ( SD.
pairs, using the computer program Statgraphics Plus for Windows 4.0.
Results and Discussion Chelate Induced Pb Plant Uptake. Because Pb has limited bioavailability in the soil, solubilizing the Pb in the soil by addition of chelates and facilitating its transport to the shoots of plants is vital to the success of phytoextraction. Cabbage (Brassica rapa) was selected as a test plant for phytoextraction study in a laboratory conditions due to its substantial Pb phytoextraction potential (17) and ability to grow in limited space available in soil columns. After harvesting, the dry biomass yield of cabbage varied between 2.11 and 7.06 g, irrespective of soil treatment. In all treatments where EDTA and EDDS were applied, visual symptoms (necrotic lesions on the leaves) of Pb or chelate toxicity were observed, followed by a rapid senescence and drying of the plant shoots. The analysis of plant material indicated that the addition of both tested chelates, EDTA and EDDS, to the soil increased the concentrations of Pb in the leaves of the test plant compared to the control. In EDTA treatments, plant tissue Pb concentrations reached 817.3 ( 106.7 mg/kg of dry biomass, in the EDDS treatment 463.4 ( 112.3 mg/kg and in the control 5.2 ( 2.6 mg/kg. These results confirm the not yet surpassed efficiency of EDTA for Pb phytoextraction compared to other chelates (13). EDDS has a strong chemical affinity for Pb (complex stability constant, Log Ks ) 12.7) but still lower than EDTA (Log Ks ) 17.88) (19). The lower Pb phytoextraction efficiency of EDDS compared to EDTA is also presumably governed by its biodegradability. In treatments with EDTA and EDDS amended soil, Pb concentrations in plant material were 157 and 89 times higher compared to control treatments. However, the mass balance of Pb extracted from the soil showed that even in EDTA treatments, the percentage of Pb phytoextracted in one phytoextraction cycle was only 0.05 ( 0.02% of the total Pb present in the soil (Table 2). This is far from the Pb plant concentrations required for efficient soil remediation within a reasonable time span. Plant tissue Pb concentrations exceeding 1% of dry biomass would be required to reduce
soil Pb concentrations by 500 mg/kg over 20-25 years using plants with a high biomass yield (20 tons/ha of dry matter). Cabbage was suitable plant for laboratory phytoextraction experiments in soil columns. However, due to insufficient Pb accumulating capacity and low biomass, cabbage cannot be regarded as candidate plant for phytoextraction in a field scale. Our results on Pb plant uptake are comparable with experiments by Epstein et al. (25), in which the addition of 5 and 10 mmol EDTA/kg of Pb contaminated soil increased the shoot Pb concentrations in Brassica juncea 26-fold and more than 100-fold, respectively. However, other authors have reported much higher plant Pb concentrations induced by chelate application. The possible reasons for higher Pb concentrations reported by Blaylock et al. (9) and Huang et al. (13) compared to our results are as follows: the harvest of very young plants, favorable speciation of heavy metals in soil, and an experimental set up in which no losses of the phytoavailable Pb in the chelate-Pb complex are due to leaching off the substrate. In a field lysimeter study, Huang et al. (13) measured 785 mg/kg (28-fold increase) of Pb in Brassica juncea grown in 5 mmol EDTA/kg soil treatment. The efficiency of chelate induced phytoextraction of Pb seemed therefore critically dependent on the experimental set up, the environmental conditions, and the soil type. The field trials of Huang et al. (13), the results of Epstein at al. (25), and our results suggest that the rate of Pb removal using conventional plants and growth conditions is insufficient. The introduction of novel traits into high biomass plants in a transgenic approach is a promising strategy for the development of effective phytoextraction technologies. A number of transgenic plants have been generated in an attempt to modify the tolerance, uptake, or homeostasis of trace elements (26). Among new approaches, transgenic plants specifically tailored for the phytoextraction of heavy metals have recently been used. Transgenic tobacco accumulated 90% more of Cd in above-ground parts then control plants, without visible differences in the growth characteristics (27). Permeable Reactive Barrier. The construction and reactive materials in our permeable barriers were a layer of nutrient enriched substrate with a high water sorption capacity and with extensive surfaces on which microbial films could form and a layer of apatite mixed with soil for precipitation of heavy metals. The main purpose of the substrate layer of the permeable barrier was enhanced microbial degradation of the Pb-EDDS complex and subsequent binding of released Pb into the solid phase of the soil or substrate. Once Pb ions are released from the chelatemetal complex, they can bind to the solid phase, become immobilized, and are no longer subject to leaching. We induced the formation of microbial activity within the substrate layer by providing nutrients to carrier substrate materials: vermiculite, hydrogel, or peat. The evolution of microbial activity in substrate layers was followed in insulated soil columns as a temperature increase due to metabolic heat. As shown in Figure 1, the temperature in the substrate layers in columns with permeable barriers reached a peak between days 2 and 3 and remained higher than in the control column approximately until day 10. During the time of increased metabolic heat, microbial colonization of the substrate layer presumably occurred. The carrier materials for the substrate layer of permeable barrier were chosen on the basis of their high water sorption capacity. This was 81.8 ( 2.7, 99.2 ( 0.0, and 87.8 ( 4.0% for vermiculite, hydrogel, and peat, respectively. Retention of enough water in the substrate layer was important for microbial colonization and survival. Furthermore, a high water sorption capacity was important to retain the soil solution with Pb-chelate complex in the substrate layer and
FIGURE 1. Difference in temperature in soil of control column and temperature in peat, hydrogel, and vermiculite layer of columns with horizontal barrier. Temperature increase was generated as metabolic heat. thus to prolong the time available for microbial degradation of Pb-EDDS and for Pb binding and in this way to increase the efficiency of the permeable barrier. Biodegradabilty of chelate is essential for this type of permeable barrier to function. Other than EDDS, nitrilotriacetic acid (NTA) is the only other biodegradable chelate already tested for phytoextraction. However, NTA has a lower chemical affinity for Pb than EDDS (19) and is not a naturally occurring substance. Chelate Induced Pb Leaching. Leachates from the soil columns were collected and analyzed for Pb content. In columns with no chelate addition and no permeable barrier installed, the concentrations of Pb in leachates were under the detection limit of the instrument (
