Article pubs.acs.org/est
Source Identification and Sedimentary Record of Polycyclic Aromatic Hydrocarbons in Lake Bled (NW Slovenia) Using Stable Carbon Isotopes Marinka Gams Petrišič,†,‡ Gregor Muri,§ and Nives Ogrinc*,†,‡ †
Department of Environmental Sciences, Jožef Stefan Institute, Jamova 39, 1000 Ljubljana, Slovenia Jožef Stefan International Postgraduate School, Jamova 39, 1000 Ljubljana, Slovenia § Institute of Public Health Kranj, Gosposvetska 12, 4000 Kranj, Slovenia ‡
S Supporting Information *
ABSTRACT: A combination of molecular and stable isotope analyses was used to trace and identify the sources of polycyclic aromatic hydrocarbons (PAH) in sediments of Lake Bled (NW Slovenia). Sediment samples were taken from two locations with contrasting depositional regimes: Zaka Bay, with permanently oxic bottom and station D, where anoxic conditions prevail throughout the year. The concentrations of PAH in surface sediments at the two locations were comparable and higher than in previous studies, reaching 4230 and 4380 ng g−1, respectively. It was found that retene (Re) and perylene (Per) are both mainly of natural origin in Zaka Bay while, at station D, the value of δ13C determined at a depth of 12−14 cm in the 1950s indicated that Re was of pyrolytic origin. The distribution of δ13C values of other individual PAH showed that PAH input to lake sediments was of pyrolytic origin, likely dominated by coal and later in 1950s also by wood burning. PAH from vehicular emissions could also contribute to the overall isotope signatures at the depth of 12−14 cm at station D and Zaka Bay corresponding to the period 1953−1961.
■
combustion, 8−12 but studies of stable carbon isotope composition of PAH in sediment cores are limited.13,14 Studies in aquatic systems focused mainly on surface sediments.4,6,7 McRae et al.14 have shown that PAH from other anthropogenic sources in urban lakes are more enriched in 13C than combustion-derived PAH. Various sources of PAH have also been distinguished in sediments along the St. Lawrence River.6 Localized 13C-enrichment in the stable isotope signature of PAH was shown to be due to petroleum-related and aluminum smelter contributions against the regional backdrop of combustion-dominated PAH sources. We have used a combined molecular and isotopic approach to trace and identify the sources of PAH in lacustrine sediments of Lake Bled (NW Slovenia). In addition, traditional methods of identifying contaminant sources by compositional information have been compared with the stable carbon isotope composition approach to show the utility of both approaches for source apportionment for past deposition of PAH in sediment.
INTRODUCTION Polycyclic aromatic hydrocarbons (PAH) constitute a large group of Persistent Organic Pollutants (POP) containing from two to six fused benzene rings. Certain PAH are among the most carcinogenic substances known and can be acutely toxic or genotoxic depending on the number and configuration of the benzene rings and the presence and position of their substituents.1 PAH exhibit different molecular distribution according to their origin formed during incomplete organic matter combustion (pyrolytic origin) or natural and anthropogenic fossil fuel combustibles (petrogenic origin). Source apportionment techniques are of particular interest motivated by environmental remediation, prevention of future contamination, and evidence in litigation. Compound-specific isotope analysis (CSIA) is a useful tool for studying the carbon cycle on the molecular scale and is important in identifying sources of PAH. This is possible because compound specific isotope composition does not appear to be modified by chemical, physical, or biological changes, as are the traditional molecular compositions.2,3 Recently, this technique was successful in identifying sources of PAH in many environmental systems including sediments.2,4−7 The sedimentary records of PAH concentrations have frequently been studied to develop historical records of © 2013 American Chemical Society
Received: Revised: Accepted: Published: 1280
September 21, 2012 December 26, 2012 January 3, 2013 January 3, 2013 dx.doi.org/10.1021/es303832v | Environ. Sci. Technol. 2013, 47, 1280−1286
Environmental Science & Technology
Article
Figure 1. Map showing the location of Lake Bled, NW Slovenia with sampling sites Zaka Bay and station D in the West Basin.
■
EXPERIMENTAL SECTION Study site description. Lake Bled is an urban subalpine lake located in NW Slovenia (Figure 1) having a surface area of 1.44 km2 and an elevation of 475 m above sea level. The lake is divided into two basins: the deepest, western basin with a depth of 30.5 m and the eastern basin with a depth of 24 m. The surficial inflows are two small streams, Mišca and Solznik, in the western basin, whereas water outflow proceeds through the Jezernica stream into the River Sava. The lake is stratified with an anoxic hypolimnion for most of the year, except during early spring and late autumn. In the shallower regions, oxidizing conditions prevail throughout the water column. Lake Bled is located in an industrially, touristically, and agriculturally developed area. To improve the quality of lake water two amelioration projects have been undertaken: freshwater inflow, diverted in the early 1960s from the River Radovna, and pumping of anoxic water from the western basin into the Jezernica. The residence time in Lake Bled was shortened from 4 years to 1.1 year after the amelioration projects took place. According to OECD15 criteria, the lake is now classified as mesotrophic. Surficial sediment of Lake Bled is composed of dark gyttya type clayey silt with 5%−10% of organic matter. The sediment below is fine laminated and composed of homogeneous silt and clayey silt. Mineralogically, low-Mg calcite prevails, followed by dolomite, quartz, partially of diatomaceous origin, and feldspar. Clay minerals are composed of muscovite/Illite and chlorite. Authigenic minerals are pyrite and low-Mg calcite deposited as lake chalk in restricted area south of Zaka Bay.16,17 Sampling. Samples of sediments were collected in November 2011 at two locations at the western part of the lake − in the shallow Zaka Bay in the thermocline at a depth of 12 m, and in the anoxic hypolimnion at Station D in the deepest part of the lake, at a depth of 30 m (Figure 1). In Zaka Bay the water column is supersaturated with oxygen down to the bottom during most of the year. Undisturbed sediment cores to a depth of 40 cm were taken with a gravity core sampler equipped with a Plexiglas tube (6 cm i.d.). Sediment samples were extruded, cut, and stored in glass jars. All utensils and glassware were prerinsed with distilled water and organic solvents to prevent contamination. In laboratory sediment
samples were freeze-dried, homogenized, and stored frozen until analyzed. Analyses. All solvents and reagents (JT Baker, Mallinckrodt Baker) were of analytical grade or higher. Dry sediments were weighed into cellulose extraction thimbles and spiked with deuterated PAH (naphthalene-d8, acenaphthene-d10, phenanthrene-d10, chrysene-d12, and perylene-d12; Dr. Ehrenstorfer) for the correction of the mass spectrometer. This step was followed by the extraction procedure described in detailed by Muri and Wakeham.18 After Soxhlet extraction with dichloromethane, extracts were concentrated by rotary evaporator, solventexchanged to hexane, and fractionated on a glass silicagel column. After eluting aliphatic hydrocarbon with hexane, PAH were eluted using additional 25 mL of hexane and 20 mL of hexane/toluene (3:1). Both PAH fractions were combined and concentrated with a rotary evaporator. Fractions were dried using nitrogen and redissolved in iso-octane for instrumental analysis. Eighteen PAH compounds, that is acenaphthene, acenaphthylene, phenanthrene, anthracene, fluoranthene, fluorene, pyrene, benzo[a]anthracene, chrysene, benzofluoranthenes [b and k isomers], benzo[a]pyrene, indeno[1,2,3cd]pyrene, benzo[ghi]perylene, dibenzo[a,h]anthracenes, napthalene, and 1-methyl and 2-methyl naphthalene were determined in the samples. The individual compounds were analyzed and identified with an Agilent Technologies 6890 GC coupled to a MSD using a DB-5 ms (30 m × 0.25 mm i.d., 0.25 μm) fused phenylmethylpolysiloxane capillary column. The PAH were determined in selected ion−monitoring mode with an ionization energy of 70 eV, whereas their identification was based on the m/z peaks corresponding to the molecular weights of the individual PAH. Concentrations of PAH were calculated using response factors; the peak area of the individual PAH was normalized with the peak area of the corresponding deuterated surrogate. In addition, procedural blanks were run with each set of six samples to monitor background contamination. Compound-specific stable isotope ratios were determined using a gas chromatography-isotope ratio mass spectrometer, which are connected together by a combustion furnace (GC-CIRMS; Isoprime GV, UK). The gas chromatograph was fitted with a DB 1 ms capillary column (60 m, 0.32 mm i.d., 0.25 μm) fused by 100% dimethylpolysiloxane, using helium as carrier 1281
dx.doi.org/10.1021/es303832v | Environ. Sci. Technol. 2013, 47, 1280−1286
Environmental Science & Technology
Article
gas. Carbon isotope measurements were reported in delta notation (δ) relative to the Vienna-Pee Dee Belemnite (VPDB) standard as follows: δ13C(‰) = [(13C/12 C)sample /(13C/12 C)standard − 1]× 103
δ13C values were calibrated against reference CO2, which was introduced directly into the source two times at the beginning and end of every GC isotope determination. The precision and accuracy of δ13C values from GC-C-IRMS were determined using individual PAH compounds initially measured with an elemental analyzer (EA) coupled to IRMS (Europa Scientific 20−20 with ANCA-SL solid−liquid preparation module). The isotope values of EA-IRMS were taken as a true values of the standards and compared with GC-C-IRMS measurements. In addition, multiple compound-specific isotope measurements were performed on a standard mixture of PAH. It was found that the precision of the GC-C-IRMS measurements ranged between 0.3 and 0.5‰, whereas accuracy ranged between 0.2 and 1.0‰. Samples dissolved in isooctane were injected and stable carbon isotope composition determined by duplicate analyses with a precision of 0.3‰ for well separated PAH compounds and up to 1.0‰ for some high molecular weight coeluting isomers. Weight percentages of organic carbon (% OC) in sediments were determined following acidification with 1 M HCl using a Carlo Erba model 1108 CHNS analyzer at a combustion temperature of 1020 °C with precision expressed in terms of relative standard deviation ±2%. This analysis was performed at Marine Biological Station, National Institute of Biology, Piran, Slovenia. Isotope compositions of sedimentary organic carbon (δ13COC) were measured with a Europa Scientific 20−20 IRMS with an ANCA-SL preparation module for solid and liquid samples. Results were calibrated against reference materials: National Bureau of Standards 22 (NBS 22, oil) and International Atomic Energy Agency standard for carbon and hydrogen (IAEA-CH-7). The precision of measurements was ±0.2‰. Sediments had previously been dated radiometrically using a γ-ray spectrometer equipped with a high-purity Ge well-type detector. 210Pb and 137Cs activities in each sediment slice were measured. Sedimentation rate was obtained based on unsupported 210Pb activities levels using the constant rate of supply (CRS) model. The dating was additionally confirmed by 137 Cs. An average sedimentation rate of 4.5 mm yr−1 was used for age assessment at Zaka Bay18 and of 2.4 mm yr−1at station D.17
Figure 2. Depth profiles of organic carbon content (OC), δ13C values, and concentrations of total polycyclic aromatic hydrocarbons (PAH) from the sediments in Zaka Bay and Station D. δ13C values for retene (Re) are also included. Depths at which δ13C values of PAH were determined are marked by gray color.
depth. The maximum concentrations up to 7650 ng g−1 were reached at depths of 12−14 and 14−16 cm. These depths had been dated previously as about 1944−1961.18 The concentrations were lower below 16 cm, being 2470 ng g−1 at 25 cm (deposited around 1907). Concentrations of individual PAH species ranged from 26 to 250 ng g−1, whereas at the depths of 12−16 cm where the maximum of total PAH concentrations was observed they ranged from 87 ng g−1 to 703 ng g−1 (Table S1 of the Supporting Information). Most PAH, that is fluoranthene (Fl), pyrene (Py), benzo[a]anthracene (BaA), chrysene (Chy), benzo[e]pyrene (BePy), benzo[a]pyrene (BaPy), indeno[1,2,3-cd]pyrene (IPy), and benzo[ghi]perylene (BghiPer) exhibited a similar depth profile, although their abundances differed. Perylene (Per) and retene (Re) had different depth patterns. Concentrations of Per were quite uniform in the first 10 cm, with an average concentration of 53 ng g−1, but reached 324 ng g−1 deeper in the sediments. Re concentrations were low at the surface, but slightly higher, up to 87 ng g−1, at depths corresponding to the period from 1953 to 1961. The concentrations again decreased deeper in the sediment to 8 ng g−1. The depth profile of total PAH concentrations in Zaka Bay differed significantly from that at sampling station D (Figure 2). In addition, the deepest sample
■
RESULTS AND DISCUSSION Organic carbon contents in sediments at station D ranged from 3.6 to 5.8 wt.% (average = 4.1 ± 0.4%, n = 18), the highest value being at a layer dated to 1902−1911 (24−26 cm). δ13COC values ranged from −34.8 to −31.5 ‰ (Figure 2). Such low δ13C values are often found in eutrophic lakes with an anoxic hypolimnion and in strongly reductive sediments.19−21 The OC content in Zaka Bay ranged from 3.5 to 4.8 wt.%. Values of δ13COC ranged from −31.3 to −29.4‰ and were higher than those at station D. These results indicate a greater influence of terrestrial OC at this station (Figure 2). Sedimentary total PAH concentrations were measured at both sampling locations, together with the concentrations and isotope composition of OC (Figure 2). Concentrations at the surface at station D were 4380 ng g−1 and decreased with 1282
dx.doi.org/10.1021/es303832v | Environ. Sci. Technol. 2013, 47, 1280−1286
Environmental Science & Technology
Article
insufficient for determination of its isotope composition at station D. However, δ13C values for Re were determined at three depths at the surface (0−1, 2−3, and 3−4 cm) and at a depth of 12−14 cm. The values in the upper part of the sediment ranged from −31.3‰ to −27.8‰, whereas a value of −25.1‰ was recorded deeper in the sediment (Table S1 of the Supporting Information). δ13C values in PAH determined at the surface and a depth of 24−26 cm in Zaka Bay ranged from −32.2‰ to −21.2‰ (Table S2 of the Supporting Information). The lowest δ13C value was recorded for Re and the highest for Fl near the surface. Values of δ13C for Re ranged from −32.2‰ to −28.2‰ and for Per from −29.3‰ to −26.1‰. Per and Re can have both anthropogenic and natural origins. Ret has been proposed as a tracer for the combustion of coniferous wood.25 It is a derivative of abietic acid present in coniferous resins26,27 and is present in extracts of algal and bacterial organic matter.28 The concentration depth profile of Ret at station D follows those of the pyrolytic PAHs suggesting that Re at this location probably originated from a pyrolytic source such as coal-burning emissions. This suggestion is supported by the δ13C value of −25.1‰, determined at a depth of 12−14 cm for Re, which falls within the range of δ13C values of pyrogenic PAHs (Table S1 of the Supporting Information). However, in the upper part of the sediment, lower δ13C values of Re (from −31.3‰ to −27.8‰) indicate a predominantly natural origin (Figure 2). The δ13C values of Re in Zaka Bay, ranging from −32.2‰ to −28.2‰, coincide with the bulk δ13COC values in sediments (Figure 2) suggesting that Re can originate from similar organic precursors and has a mainly diagenetic origin. At this location, the percent of allochthonous organic material, which is delivered by the small stream Solznik, is higher than that at station D. The stream Solznik drains the partially coniferous Lake Bled watershed. The origin of Per is less specific and has been the subject of considerable debate. Emissions from automobiles29 and municipal incinerators30 and in situ diagenesis of marine and terrestrial organic matter31 have been cited as potential sources of Per. The concentration profiles of Per in Lake Bled sediments differ markedly from those of the pyrogenic PAHs. Postdepositional, in situ diagenetic formation has been suggested as explaining the increase of Per with depth at station D.22 This assumption was further supported by Py/Per ratios ∼1.0 found in the older western basin sediments. The Py/Per for diagenesis was suggested to be
