Research Sorption of Lipophilic Organic Compounds to Wood and Implications for Their Environmental Fate S T E F A N T R A P P , * ,† KARINA S. B. MIGLIORANZA,‡ AND HANS MOSBÆK† Miljø & Ressourcer, Technical University of Denmark, DK-2800 Lyngby, Denmark, and Laboratory of Ecotoxicology. Faculty of Natural Sciences, Mar del Plata University, Funes 3350, 7600 Mar del Plata, Argentina
TABLE 1. Biomass Distribution in a Central European Quercus-Carpinus Deciduous Foresta compartment green plants leaves of trees roots branches branch growth per year stems stem growth per year animals birds and mammals soil animals incl. earth worms soil microflora sum a
The sorption from water to wood (KWood) of 10 organic chemicals (log KOW, 1.48-6.20) was experimentally determined for oak (Quercus robur) and basket willow (Salix viminalis). Linear regression yielded log KWood ) -0.27 (( 0.25) + 0.632 (( 0.063)log KOW for oak (r ) 0.90, n ) 27) and log KWood ) -0.28 (( 0.40) + 0.668 (( 0.103)log KOW for willow (r ) 0.79, n ) 27). According to an equilibrium-partitioning model, wood should be an important storage compartment for lipophilic environmental chemicals, but this is contrary to analytical results. Diffusive uptake from air into wood was estimated to be a relevant transport process only for chemicals with a high KAW. Uptake of chemicals from soil via xylem into stem was simulated with a dynamic onecompartment model. This pathway seems to be important for chemicals with low and intermediate lipophilicity. In large trees, the chemicals are retained for a long time. If metabolism inside the stem occurs, wood can serve as a “safe sink” for environmental chemicals. This might be of use in phytoremediation.
Introduction Forests are the dominating ecosystems of our planet. In a Central European forest ecosystem, woody stems and branches provide most of the biomass. Although the figures in Table 1 are not complete (the value for roots is an estimate, it lacks insects, and the value for soil microflora is questionable), roughly 95% of forest biomass is stems and branches. The large pool of biomass probably acts as a storage or a sink compartment for environmental chemicals. Meredith and Hites (1) found PCBs in the bark of black walnut and tulip poplar trees, which were exposed to a PCB-contaminated landfill, and in the bark of white oak trees 14 km away. Concentrations were higher in the outer bark. There was little evidence that PCB was present in the wood, values were near the blanks and at least a factor 62 smaller than in the outer bark of walnut (sum of 14 PCB congeners). Simonich and Hites (2) calculated a mass balance of polycyclic aromatic hydrocarbons (PAH) in the atmosphere-vegetation system. According to their findings, bark has almost as high average * Corresponding author e-mail:
[email protected]; telephone: +45 4525 1622; fax: +45 4593 2850. † Technical University of Denmark. ‡ Mar del Plata University. 10.1021/es000204f CCC: $20.00 Published on Web 03/07/2001
2001 American Chemical Society
biomass (t/ha) 4 10 30 2.5 240 2.5 0.0038 0.8 0.3 285.1
% 1.4 3.5 10.5 84.2 0.001 0.28 0.1
Dry weight per ha in summer, ref 29.
PAH concentrations as needles based on dry weight mass. On an area basis, bark contains more PAH than leaves, needles, and seeds together (wood was not investigated). Simonich and Hites concluded that “vegetation is a major pathway through which lipophilic organic compounds are removed from atmosphere” (2). In a follow-up study, the sorption capacity of bark was used to evaluate the global environmental distribution of 22 potentially harmful organochlorines (3). In a critical review on organic pollutant accumulation in vegetation, the wood compartment was not considered (4). Schramm et al. (5) simulated concentrations in trees with the UNITTree model. They found concentrations in wood comparable to those in soil and needles of the trees. Wood was assumed to sorb as strong as organic carbon in soil. No comparison to measured values was given. Mackay and Gschwend (6) recently measured sorption of monoaromatics (benzene, toluene, o-xylene) to wood of Ponderosa pine and Douglas fir. Partition coefficients wood to water (Kwood) were between 6.6 and 28 (mg/g of dry wood to mg/ mL of water). This indicates a rather high sorption capacity of wood. We therefore set up a series of experiments to confirm these findings and extend them to more lipophilic compounds. We also give the equilibrium distribution for an oak forest and make some estimates about the dynamics of uptake into stems.
Materials and Methods Wood. The 2-year-old branches of basket willow (Salix viminalis) were provided from Aage Bach, Tylstrup, Denmark. The 3-5-year-old branches of common oak (Quercus robur) were cut from the university campus in Lyngby, Denmark. The branches had a diameter of 1-2 cm. They were peeled and filed with a rough file to give fine splints of about 1-2 mm length with a diameter of less than 0.2 mm. Dry weight, measured by drying the splints overnight at 105 °C, was 60.8% of wet weight for oak (SD ) 1.7%, n ) 3) and 68.4% for willow (SD ) 2.0%, n ) 3). Chemicals. A stock solution of 10 chemicals (provided from Sigma, Denmark) with individual masses listed in Table 2 was prepared. Approximately 50 mg of the mixture was given into 500 mL of distilled water for each experiment. Experimental. The solution was whirled for 12 h to ensure mixing. Wood splints and solution were mixed in different ratios (1:100 or 1:10, wood:water, weight based) in glass vessels and shaken for 12 h. After that, samples were centrifuged at VOL. 35, NO. 8, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
1561
TABLE 2. Chemicals and Physicochemical Propertiesa compound
log KOW
KAW
phenol benzene chlorobenzene o-xylene naphthalene 1,2-DCB lindane 1,3,5-TCB dieldrin DDT
1.48 2.13 2.78 3.16 3.35 3.40 3.76 4.02 5.14 6.20
2.2 × 10-5 0.23 0.15 0.22 0.023 0.1 4.5 × 10-5 0.15 0.00046 0.0011
solubility (mg/L)
stock (mg)
92 000 1769 463 175 30 130 6.3 6.59 0.2 0.0034
49.39 19.73 9.93 20.02 4.85 11.25 0.86 1.39 1.11 1.47
a From ref 13, except dieldrin and trichlorobenzene, which were from ref 30. DCB is dichlorobenzene. TCB is trichlorobenzene. Sum stock is 120 mg. KOW is the partition coefficient between n-octanol and water. KAW is the partition coefficient between air and water.
2000 rpm for 20 min. Experiments with a 1:10 ratio had to be canceled at that step because it was impossible to separate wood from water. Analytical. From the 1:100 treatments, aqueous samples were taken and extracted with a pentane-diethyl ether (15:85) mixture for 2 h. Bromonaphthalene was used as an internal standard. Samples were manually injected into a HP 5890 series II gas chromatograph with a 30 m HP-1 column (diameter 0.53 mm; 1 µm layer) and detected using parallel connected FID and ECD. The concentrations of compounds were checked against a certified standard (Supelco Denmark). The detection limits with 10 mL water-sample were
