Environ. Sci. Technol. 2007, 41, 82-87

deposited during the peak period of combustion. The Elemental and Molecular Evidence temporal trends of soot-BC observed in our lake cores do of Soot- and -Derived Black not agree with published historical reconstructions Carbon Inputs to New York City’s based on fuel consumption and estimated emission factors. Atmosphere during the 20th Century Introduction P A T R I C K L O U C H O U A R N , * , † , ‡ Black carbon (BC) is a generic term that was initially S T E V E N N . C H I L L R U D , ‡ introduced by Novakov (1) to illustrate the role of soot in S T E P H A N E H O U E L , ‡ , # B E I Z H A N Y A N , § atmospheric sulfur chemistry. Goldberg (2) further charac- D A M O N C H A K Y , ‡ C O R N E L I A R U M P E L , | terized BC based on particle size and formation conditions C L A U D E L A R G E A U , ⊥ G E R A R D B A R D O U X , | as a spectrum of highly recalcitrant organic residues re- D A N W A L S H , ‡ A N D R I C H A R D F . B O P P § maining after organic matter combustion (whether biomass Department of Earth and Environmental Sciences, Columbia or ). With time, the BC definition has evolved into University, Geoscience Building, Room 110, P.O. Box 1000, a continuum model which represents combustion byproducts 61 Route 9, West Palisades, New York 10964-8000, ranging from slightly charred materials, retaining original Lamont-Doherty Earth Observatory of Columbia University, structural information of the parent material, to highly Palisades, New York 10964, Earth and Environmental condensed refractory soot (3, 4). Simple particle size dif- Sciences, Rensselaer Polytechnic Institute, Troy, New York ferentiation between char and soot has been augmented with 12180, Laboratoire Bioge´ochimie et EÄ cologie des Milieux methods employing optical, physical, chemical, and/or Continentaux, Site du Centre INRA Versailles-Grignon, thermal treatments to separate noncombustion organic Baˆtiment Eger, 78850 Thivernal-Grignon, France, and carbon (OC) from BC. Although BC is often referred to as Laboratoire de Chimie Bioorganique et Organique Physique, being primarily elemental carbon in the form of , it UMR CNRS 7618 BIOEMCO, EÄ cole Nationale Supe´rieure de is seldom pure and normally includes varying proportions Chimie de Paris, France of other atoms (5). Urban centers are major sources of combustion-derived particulate BC and OC to the atmosphere, and their expansion in the 20th Century has led to environmental impacts of BC Soot black carbon (here expressed as GBC) is present in that range from local health effects on humans (6, 7) to sediments of Central Park and Prospect Park Lakes, New potentially global influence on the earth’s radiation budget York City (NYC), and peaks in the middle of the 20th (5, 8). The energy and transportation sectors have been Century at the highest values (1-3% dry weight) ever recognized as major emitters of fine carbonaceous particles, reported in urban lakes. During that period ( 1940-1970), or soot-BC, which can redistribute over vast areas (9). the GBC represents up to 28% of the total organic∼ Researchers have attempted to reconstruct temporal trends carbon (OC). Radionuclide-normalized whole core inventories of soot-BC using historical data on fuel consumption and of accumulated GBC are similar in the two lakes which estimated fuel-specific BC emission factors (9-11). However, are separated by 15 km, suggesting that emissions of fine emission factors are poorly constrained, resulting in the soot particles may∼have accumulated homogeneously recognition of the need for additional reconstruction ap- proaches (5). Ice cores provide the potential for obtaining over at least the urban center of NYC. The distribution of the history of long-range transport of fine-particles of BC polycyclic aromatic hydrocarbons (PAHs) in the sediments emitted from multiple source regions over time. Urban lake is decoupled from that of GBC. The highest levels of total cores, on the other hand, may record the history of BC and PAHs correspond to peak use for space heating in other combustion-associated contaminants, including PAHs NYC in the early 1900s. In contrast, GBC concentrations were and metals, near their sources (12-16). They can provide a highest in the mid 1900s, a period when oil combustion chronology of deposition of combustion-derived carbon- dominated local fossil fuel use and incineration of municipal aceous aerosols, over the past 100-150 years, a period of solid waste (MSW) was common practice in NYC. accelerated industrialization characterized by major changes Decreases in GBC levels observed in more recently in combustion technologies and fuel use. Furthermore, deposited sediments are consistent with improvements in reconstructing atmospheric deposition of BC in urban lakes can constrain historical exposure to combustion-derived particle emissions control systems. Non-soot BC (char) aerosols for health studies of urban populations. was identified by a high carbon to nitrogen (C/N) ratio that This study uses two lakes within the airshed of NYC to persisted after correction for GBC. This likely tracer of establish the history of BC emissions since the end of the MSW incineration was estimated to contribute an additional 19th century and focuses on two components of BC: 35% of total organic carbon found in the sediments thermally recalcitrant soot-BC and thermally labile char ∼ residues derived from inefficient combustion processes, such * Corresponding author phone: (409) 740-4710, fax: (409) 740- as municipal solid waste (MSW) incineration. 4787, e-mail: [email protected]. Present address: Depts. of Ocean- ography and Marine Sciences, Texas A&M University, 5007 Avenue U, Galveston, Texas 77551. Materials and Methods † Columbia University. Push cores were collected in 1996 from Central Park Lake ‡ Lamont-Doherty Earth Observatory of Columbia University. (CPF) and in 2002 from Prospect Park Lake (PPL7), located § Rensselaer Polytechnic Institute. respectively, in NYC’s boroughs of Manhattan and Brooklyn. | Site du Centre INRA Versailles-Grignon. ⊥ UMR CNRS 7618 BIOEMCO. The lake history and chronology of contaminant inputs to # Present address: University of Colorado at Boulder, Department Central Park Lake have been described previously (12, 16). of Chemistry and Biochemistry, 215 UCB, Boulder, CO 80309. We are not aware of any previous study describing the

82 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 1, 2007 10.1021/es061304+ CCC: $37.00 © 2007 American Chemical Society Published on Web 11/29/2006 TABLE 1. Organic Carbon (OC), Nitrogen (N), Organic Carbon/Nitrogen Atomic Ratio ((C/N)a), and Black Carbon (CTO-BC and GBC) Values for Four Reference Materials (2 Soot Samples and 2 Estuarine Sediments)a OC N CTO-BC GBC BC (Lit) (%dw) (%dw) (C/N)a (%dw) (%dw) (%dw) diesel soot (SRM 2975) 86.5 ( 1.7 0.42 ( 0.09 240 ( 52 64.4 ( 0.8 63.7 ( 8.7 63 ( 4.1c-68.2 ( 0.9d n-hex soot (U. of Denver) 92.8 ( 1.0 bdb 43.8 ( 2.3 42.2 ( 1.5 41.1 ( 0.8e-44 ( 5.3c NY/NJ sediment (SRM 1944) 4.4 ( 0.4 0.22 ( 0.01 23 ( 2 0.96 ( 0.04 0.82 ( 0.07 0.66 ( 0.16d-0.80 ( 0.02d Balt Harbor (SRM 1941b) 3.1 ( 0.1 0.26 ( 0.01 14 ( 1 0.41 ( 0.08 0.41 ( 0.04 0.51 ( 0.14e-0.58 ( 0.05d a CTO-BC: Chemothermal oxidation method (Gustafsson et al. (21)); GBC: chemothermal oxidation method with prior demineralization and hydrolysis (Gelinas et al. (22)); BC Lit: literature values for soot BC in the reference materials listed. b bd: below detection level. c Nguyen et al. (23). d Gustafsson et al. (21). e Hammes et al. (4). environmental geochemistry of PP Lake sediments. Both lakes exact soot-BC standards exist at present (4, 20), accuracy were constructed in the 1860s as part of the 19th Century can only be assessed by comparing results from similar Park works in Metropolitan NYC. They have surface areas of treatments of selected reference materials. Soot-BC was approximately 7.1 104 m2 and 2.4 105 m2, respectively, measured on several BC-containing NIST standard reference × × and a drainage to surface area ratio of 10-15 within a heavily materials using both the GBC and CTO methods and the ∼ urbanized environment (12). Both lakes are shallow with a results were compared to published data (Table 1). Based on deepest water column of 1.2 and 1.8 m where CPF and PPL7 replicate analyses (n ) 4) of each sample, the precision of were collected, respectively. the GBC method averages 9% (range 4-14%). These data Upon recovery, core CPF was sliced into 2 cm sections also demonstrate that both methods produce results that over its entire length (52 cm). PPL7 was sliced into 1 cm are consistent with values previously reported in the litera- sections from 0 to 10 cm, 2 cm sections from 10 to 30 cm, ture. Furthermore, comparison between GBC and CTO-BC and 4 cm slices for the remainder of the core (30-54 cm). on a series of natural sediment samples, NIST reference All sections were oven dried at 35 °C under a flow of air materials, and soot, produces a strong relationship with a filtered through a column of florisil. Dried samples were then slope of 1 over 2 orders of magnitude in soot-BC concentra- ground and homogenized with a mortar and pestle. tions (Supporting Figure 1). These results confirm that the Sediment Dating. Sediment chronologies for these cores potential for positive bias from charring of labile OM in some were determined based on 137Cs and unsupported 210Pb samples by the CTO method (22) seems to be minimized or 210 ( Pbxs) profiles as described by Chillrud et al. (12), Yan et close to negligible in samples containing high BC/OC ratios al. (16), and Chaky (17). Radionuclide activities were analyzed (24). The only apparent advantage for using the GBC method by γ spectrometry using lithium-drifted germanium and in such samples is to avoid any potential shift in stable isotopic intrinsic germanium detectors. Activities were decay- carbon values that may result from fractional charring of corrected to the sample collection date. labile organic matter. The isotopic signatures of GBC in these Elemental Analyses (Carbon and Nitrogen). Carbon and samples will be discussed in an additional paper. nitrogen analyses used either a Europa elemental analyzer (EA) (Lamont Doherty Earth Observatory) or a Carlo Erba EA Results and Discussion NA 1500 (Laboratoire de Bioge´ochimie et Ecologie des Milieux Continentaux). Organic carbon (OC) was determined on all Sediment Dating. Radionuclide depth profiles (137Cs and samples after inorganic carbon removal using HCl-fumigation 210Pb) in Central Park Lake sediments have been published (18). The average precision of elemental analyses is typically previously and are consistent with semi-continuous sediment between 2 and 5%. accumulation rates over the past century (12, 16). A similar Graphitic Black Carbon (GBC). In the recent past, several model that incorporated near-surface mixing was used to studies have investigated the potential of different methods assign depositional dates to depth sections of the Prospect to isolate reproducible and comparable amounts of BC from Park Lake core (PPL7). Section-specific radionuclide activities similar samples (4, 19, 20). These efforts reached a consensus and age assignments are provided in Supporting Information that because of the large chemical variations in BC across Tables 2 and 3. In these sedimentary environments, the 210 210 the combustion continuum, there exists no universal method unsupported Pb radionuclide profiles ( Pbxs) are con- for isolating and quantifying the total amount of BC in any sistent with the 137Cs dating model and indicate two particle sample. Instead, specific methods are now recognized for accumulation rates: a high rate ( 0.2 g cm-2 yr-1 for CPF; ∼ their potential to recover combustion material from a 0.3 g cm-2 yr-1 for PPL7) for the period 1900 to the late ∼ ∼ particular section of the continuum (3, 4). In the most recent 1960s, followed by a lower rate ( 0.1 g cm-2 yr-1 for CPF; ∼ intercomparison study, Hammes et al. (4) confirmed that 0.05 g cm-2 yr-1 for PPL7) since ca. 1970. The distinct ∼ one category, the chemothermal oxidation methods, are decrease in particle accumulation rate in both cores matches appropriate for isolating the most refractory components of a 5-8-fold decrease in airborne particle settling rate across the BC continuum, namely highly condensed structures the same period (0.047-0.076 g cm-2 yr-1 vs 0.011 g cm-2 characteristic of soot (21, 22). Other combustion-derived yr-1, respectively (25)) suggesting that lake sediments in the carbonaceous materials such as charred residues are ther- NYC urban area synchronously record changes in atmo- mally labile and thus excluded from the chemothermal spheric fallout of particles (12). In core CPF, the radionuclide analytical window (21-23). Two variations of this method profiles and their strong correspondence in sedimentation exist (Supporting Table 1). The first (CTO) (21) uses thermal rates suggest that minimum mixing occurred in this lake (12, 210 oxidation (375 °C for 24 h) to remove labile organic carbon 16). In contrast, the identical Pbxs activity in the top 5 cm ∼ followed by a simple acidification to remove carbonates. The of core PPL7 as well as the broadening of the 137Cs peak second (GBC) (22) introduces a stepwise demineralization (Figure 1) show that PPL7 sediments have undergone (1 N HCl/10% HF) followed by hydrolysis of reactive organic postdepositional mixing. The similarity of the whole core 137 210 matter (O2-free trifluoroacetic acid and HCl) prior to the inventories of Cs and Pbxs for the two cores (Supporting thermal oxidation. Information Table 4) indicates that the two coring sites have In our analyses of the lake sediment samples, we have recorded atmospheric deposition over the last century with used the GBC method (22) to quantify soot-BC. Since no similar amounts of particle focusing.

VOL. 41, NO. 1, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 83 in the two NYC lakes (SI Table 5) suggest that up to 28% of the total OC is comprised of soot-BC. Since the GBC method measures only the most thermally resistant carbon structures, this value represents a lower limit, approximately half of combustion-derived materials in these sediments (see dis- cussion below). In core CPF, the sharp increase in GBC concentration between 1930 and 1940 is co-incident with large shifts in molecular indices, such as the total concentra- tion of saturated hydrocarbons and their odd to even carbon- number ratio (SI Figure 2). These indices trace the inputs of petroleum combustion byproducts to these systems (16) and support the increased use of oil in combustion sources (space heating, vehicular traffic) in the city. In core PPL7, the lower concentrations and broader shape in the GBC peak are due mostly to sediment mixing. The total core inventory for GBC (both raw inventories and normalized either to whole-core 137 210 Cs or Pbxs inventories to account for postdepositional particle focusing (12)) in core PPL7 is similar to that of core CPF (SI Table 4) indicating that over multi-decadal time- scales, soot-BC was redistributed relatively uniformly over much of the urban airshed. The persistence of relatively high GBC concentrations since the 1980s is most likely related to increases in vehicular traffic, most importantly increased diesel combustion (11),

137 though decade-scale basin holdup for these lakes (17) and FIGURE 1. Depth profiles of Cs for cores PPL7 and CPF in Prospect sediment mixing are probably responsible for part of that Park and Central Park Lakes, respectively. Depositional ages 210 137 trend. estimated by the Pbxs model are shown for Cs peaks and deepest detectable activities in both cores. A mixing model has been applied The timing of GBC maxima in NYC lake sediments and to the Prospect Park Lake radionuclide profiles (see Supporting the finding of relatively low levels in the first several decades Information for details). of the 20th century are consistent with the records of CTO- BC in Mystic Lake in Boston. The timing of the maximum and possibly the early 20th century record of GBC deposition in Lake Washington in Seattle are also similar (15, see SI). All these urban lake results contrast dramatically with the estimated history of BC emissions in the United States derived by applying emission factors to fuel consumption data (11). That work suggests peak BC emission occurred in the 1920s- 1940s associated with coal combustion and declined by more than half by the 1980s. It is clearly important to reconcile results from these two approaches and one area of productive research would be the refinement of emission factors; the paucity of direct measurements led to the caution that “there is at least a factor of 2 uncertainty in the derived emissions” (27). Specifically with respect to coal combustion for domestic heating, the same emission factor was applied to both and and it was based on two references reporting values that differed by a factor of 4.5. Comparison of BC and PAH Trends. While BC and pyrogenic PAHs are both combustion proxies, their relative emissions can vary significantly among different fuels and among different combustion sources using the same fuel (9, 28, 29). Consequently, there is no reason to assume that these tracers should be strongly correlated in sediments of lakes with a complex history of multiple combustion-derived FIGURE 2. GBC concentrations (% dw) in Central Park Lake (core inputs. While a general covariation of BC and pyrogenic PAH CPF) and Prospect Park Lake (core PPL7). The CTO-BC values from levels has been reported in some urban lakes (15, 26), that Mystic Lake were obtained from Gustafsson et al. (26) and are is clearly not the case in Central Park Lake. In core CPF the graphed for comparison maxiumum PAH concentration occurs at 1920 (16) when ∼ Temporal Trends in Soot-BC Accumulation. Both sedi- GBC levels were close to the lowest measured in the core ment cores have very high GBC concentrations especially (Figure 3). To our knowledge, Central Park Lake is the only from the 1930s to the 1970s. The maximum levels observed lacustrine environment that has reported maximum PAH in the sediments are particularly striking, reaching 2.9% dry concentrations in 1920 rather than in the mid-1950s. Other ∼ weight (dw) in core CPF in the 1950s and averaging 1.3% dw sedimentary systems show, at most, small peaks in PAH levels in core PPL7 from the 1950s to 1970s (Figure 2). These values in early 1900s (15, 30, 31). A less pronounced decoupling of are 2 orders of magnitude higher than maximum BC levels BC and PAH levels can be seen in the Mystic Lake data (26). previously reported in most other urban and suburban lakes In the early part of the 20th century, CTO-BC concentrations ( 0.03-0.06% dw (14, 15)). Only sediments of Mystic Lake increased 3-fold while the concentration of pyrene increased (B∼oston) approach BC levels on the same order of magnitude by close to 2 orders of magnitude. as those of NYC lakes (Figure 2) with peak concentrations In the early 1900s, anthracite coal was the predominant reaching 0.7% dw in the early 1960s (26). The GBC/OC ratios fossil fuel used for space heating in NYC (32). The impact of

84 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 1, 2007 FIGURE 4. Sediment profiles of total and corrected atomic C/N FIGURE 3. GBC concentrations (% dw) and the sum of 16 pyrogenic ratios ((C/N)a and (C/N)corr) in Central Park Lake and Prospect polycyclic aromatic hydrocarbon concentrations (µg gdw-1) in Park Lake. Central Park Lake (core CPF). The values for the PAHs were obtained from Yan et al. (16). In the text, a comparison is made between the source of both soot and char-sized particles (36). While sum of 13 PAHs reported in coal combustion particles (28, 34) and maximum GBC concentrations ( 1940s to 1970s) coincide CPF sediments deposited in the 1920s; for this comparison we ∼ ∼ ∼ roughly with the timing of maximum MSW incineration (36), recalculated the sum based on the same 13 PAHs (which account more quantitative source apportionment will require ad- for 99% of the total based on 16). ditional work to determine the importance of this latter combustion source and may not be possible. In NYC during coal combustion is confirmed by a synchronous peak in this period significant amounts of fuel oil, gasoline, diesel mercury flux recorded in another Central Park Lake core fuel, and coal were also being combusted (37). For more (CPE) (32). Coal combustion is recognized to emit more PAHs recent samples, source apportionment would be further and particulates than oil and gas per unit of energy generated complicated by the need to account for the introduction and (9, 33). Data from coal soot from a commercial power plant implementation of various particle emission control tech- (11 mg PAH g BC-1; (34)) and soot from burning powdered nologies. The contribution of MSW combustion to a large anthracite and clay briquettes in cook stoves (2.1 mg PAH atmospheric flux of non-soot black carbon (char) is discussed g BC-1; (28)), can be used to obtain a first-order estimate of below. coal-combustion-derived PAH concentrations in the sedi- Elemental Indicators of Non-Soot Carbon Inputs. Total ments. For the depth interval in core CPF that accumulated OC content in the two cores increases from 2% dw, in the 1910 (the PAH maximum), the GBC concentration (2 mg first part of the 20th Century, to stable large∼values ( 13% ∼ gdw-1; Figure 3) together with the above emission factors dw) in the last decades (SI Table 5). Atomic C/N ratios∼((C/ suggest coal-combustion-derived PAH levels of 4.2-22 µg N)a; Figure 4) range from 10 to 35 and show an intriguingly gdw-1. The measured PAH concentration in that interval was high peak in the middle of the century for both cores (35 and 27 µg gdw-1 (Figure 3). Similarly, the particle flux needed 22 for CPF and PPL7, respectively). Potential sources of ∼ to account for the focusing-corrected GBC flux to these carbon-rich OM that could increase substantially the (C/N)a sediments (0.23 mg cm-2 yr-1), can be estimated from the ratios to values beyond the 10-15 commonly observed in BC loading on particulates released from anthracite com- most lakes (37, 38) include terrigenous OM and soot. bustion (0.3%; (28)). This calculation yields a particle Terrigenous OM is relatively depleted in nitrogen and large deposition of 77 mg cm-2 yr-1, which is remarkably similar inputs of this material to lake sediments often result in (C/ ∼ to the particle settling rate measured in the city prior to the N)a ratios greater than 15-20 (14). Considering, however that early 1930s ( 60-80 mg cm-2 yr-1; (25)). Consequently, it eutrophic conditions have prevailed in Central Park Lake ∼ appears that during the early 20th century, combustion of (Central Park Conservancy Annual Reports, 1903-1995) and anthracite for residential heating may have been responsible that nitrogen shows relatively good preservation potential in for a significant portion of the atmospheric particle and sediments under highly productive water columns (14, 38), associated PAH flux in this urban environment. Unfortu- it seems unlikely that accumulation of an unusually high nately, the lack of comprehensive emission factor data leads fraction of carbon-rich terrigenous organic matter is re- to significant uncertainty in our estimates (see SI). sponsible for the high (C/N)a ratios in our sediments. In NYC, the decoupling of BC and PAH deposition Because soot is composed predominantly of carbon (4) illustrates the city’s complex history of combustion and with negligible amounts of nitrogen ((C/N)a > 100; Table 1), variable emissions of combustion-derived constituents. A their incorporation in environmental matrices could also major historical component of combustion in the NYC affect the bulk (C/N)a ratios. Jeong et al. (39) have reported airshed was the incineration of MSW (35). This has been such an impact in modern speleothem layers showing a shown to be a major source of the atmospheric deposition doubling of the OM (C/N)a ratios (10 to 21) through the recent of several metals (12, 32). Data on the size distribution of addition of ultrafine C-rich soot materials. The impact of particles emitted from this source are limited (see refs in this latter input on bulk (C/N)a ratios can be corrected in Walsh et al. (36)), but in terms of mass it is clearly a significant NYC lake sediments by removing the GBC from total OC

VOL. 41, NO. 1, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 85 content and using the OCcorr value in the elemental ratio contained a high proportion of incompletely burned charred calculation. When this correction is applied (Figure 4), the residues (35). These incineration conditions resulted in (C/N)a ratios in PPL7 stabilize at 17 during the 1940-1980s emissions that included large particles ( 10 µm) as well as suggesting that a substantial fraction∼ of the carbon-rich soot (<1 µm) (refs in Walsh (36)). The large∼ particles tend to material driving the high (C/N)a ratios in this core comes deposit rapidly, near the source of emission (3). Consequently, from fine soot depositing throughout the urban airshed. The the high density of incinerators supports the 4-fold larger correction in core CPF lowers the (C/N)a ratios to values accumulation rate of incinerator particles estimated for ranging from 24 to 28 during the 1940-1970s. Concurrent Manhattan vs the rest of New York’s metropolitan area (530 decreases in lignin concentrations, a vascular plant-specific mg cm-2 vs 120 mg cm-2, respectively (36)). These estimates polymer, during the same period in core CPF (SI Figure 3), are roughly consistent with the geographical variation of suggest that inputs of terrigenous OM are not responsible observed deposition rates of airborne particles (termed for the fact that high (C/N)a ratios persist in CP Lake sediments “sootfall” but based on total mass of deposited particles) for even after correction for GBC. The minimum in carbon NYC in the mid-1930s (25) which show that settling rates of preference index of saturated hydrocarbons ( 2) during this particles were 2-3 fold larger over Manhattan than the rest time (16) is also consistent with a relatively minor∼ influence of the city. This geographical difference could account for of terrigenous OM. the higher (C/N)corr values in Central Park Lake (Manhattan)

The relationships between the GBC/OCcorr and (C/N)a sediments compared to those from Prospect Park Lake ratios in NYC’s lake sediments fit a straight line (r2 ) 0.87 (Brooklyn). and 0.71 for PPL7 and CPF, respectively) with identical (C/ Our findings suggest that an additional source should be N)a intercepts of 11 for a GBC/OCcorr value of zero (SI Figure recognized for in the BC-continuum (3, 4). The present 4). The intercepts∼of the correlation lines correspond to GBC- “paradigm” defining the spectrum of materials within the free (C/N)a ratios and are similar to the “pre-combustion” BC-continuum considers that char inputs to the environment ratios in the deepest section of both cores. The strong are mostly derived from biomass combustion whereas soot correlations indicate that, although GBC cannot account for particle inputs are derived from industrial/vehicular com- the entire elevation of (C/N)a, the additional carbon-rich bustion. We suggest here that municipal and domestic component is combustion-related. The peaks in (C/N)corr incineration of solid waste can be a significant source of ( 17 and 28 in PPL7 and CPF, respectively) are co-incident both chars, which deposit locally and probably retain some with∼ maxima∼ in sedimentary polychlorinated dibenzo-p- structural information of the parent material (the solid waste dioxins and furans measured in CPE core (1960s) and which components incinerated), and soot. have been shown to trace the increased release of chlorinated organics from MSW incineration (17). We propose that Acknowledgments incineration chars constitute the additional source of at- We gratefully acknowledge support from a variety of funding mospherically derived combustion byproducts. sources including the Earth Institute at Columbia University, MSW incineration is known to produce significant the Lamont Climate Center, the Hudson River Foundation amounts of thermally labile char. Ferrari et al. (40) estimated (A90095; 007/94P), and NIEHS grants (ES07384 and that 20-40% of total organic carbon in combustion bottom ES009089). We are also grateful for the help of the Central ashes of MSW was thermally labile at 315 °C showing that Park Conservancy and the Friends of Prospect Park, and this material is composed of non-soot chars. Aliphatic finally for the assistance of Nicole Predki in the lab. This is moieties have been quantified as representing 40-60% of Lamont-Doherty Earth Observatory Contribution No. 6967. condensates from solid waste incinerators (41) and the predominance of lower molecular weight PAHs (three- Supporting Information Available to five-ring) observed in MSW incineration (42), further Radionuclide age assignment, tables of BC nomenclature suggests little condensation of aromatic products. This summary, radionuclide information, radionuclide normal- pyrogenic material is more akin to the small clusters of two ization, elemental and GBC concentrations, supporting to five aromatic rings observed in charred residues than in figures, and literature cited in the SI. This material is available soot (43). A similarly large contribution of alkyl carbon free of charge via the Internet at http://pubs.acs.org. moieties ( 40%), most likely present as alkyl groups attached to aromati∼c rings or in straight-chain hydrocarbons, was also reported in a diesel soot sample (22). This carbon fraction Literature Cited was shown to be lost quantitatively during thermal oxidation (1) Novakov, T. The Role of Soot and Primary Oxidants in - treatment (375 C, 24 h). That would also be the case for any Atmospheric Chemistry. Sci. Total Environ. 1984, 36, 1 10. ° (2) Goldberg, E. D. Black Carbon in the Environment: Properties char produced from incineration of paper waste that closely and Distribution; John Wiley & Sons: New York, 1985. resembled its parent material. Consequently, the presence (3) Masiello, C. A. New directions in black carbon organic geochem- of such combustion-derived carbon-rich materials can istry. Mar. Chem. 2004, 92, 201-213. explain the high observed (C/N)corr ratios. A simple mass (4) Hammes, K.; Schmidt, M. W. I.; Currie, L. A.; Ball, W. P.; Nguyen, balance suggests that near the peak in (C/N) , MSW T. H.; Louchouarn, P.; Houel, S.; Gustafsson, O.; Elmquist, M.; a Cornelissen, G.; Smernik, R. J.; Skjemstad, J. O.; Masiello, C. A.; incineration chars may comprise on the order of 30% of the Song, J.; Peng, P.; Mitra, S.; Dunn, J. C.; Hatcher, P. G.; Hockaday, OC in Prospect Park Lake sediments and 45% in Central Park W. C.; Smith, D. M.; Hartkopf-Fro¨der, C.; Bo¨hmer, A.; Lu¨er, B.; Lake. The fact that trends in (C/N)a closely track those of Huebert, B. J.; Amelung, W.; Brodowski, S.; Huang, L.; Zhang, (C/N)corr (Figure 4) suggests that incineration of MSW is also W.; Gschwend, P. M.; Flores-Cervantes, D. X.; Largeau, C.; a significant source of GBC to the cores. Rouzaud, J.-N.; Rumpel, C.; Guggenberger, G.; Kaiser, K.; Rodionov, A.; Gonzalez-Vila, F. J.; Gonzalez-Perez, J. A.; de la Prior to the installation of air pollution control systems Rosa, J. M.; Manning, D. A. C.; Lo´pez-Cape´l, E.; Ding, L. (APC) and their effective maintenance, on both domestic Comparative analyses of black carbon reference materials from and municipal incinerators in the late 1960s-early 1970s, soil, water, sediment and atmosphere and environmental waste incineration was a chronic source of particles to the implications. Global Biogeochem. Cycles; submitted. atmosphere (25, 35, 36). The inefficient conditions under (5) Hansen, J.; Bond, T.; Cairns, B.; Gaeggler, H.; Liepert, B.; Novakov, which such incinerators were operated (inexperience of T.; Schichtel, B. Carbonaceous Aerosols in the Industrial Era. EOS 2005, 85, 241-248. operators, low and variable combustion temperatures, refuse (6) Smith, K. R. Fuel combustion, air pollution exposure, and charging during combustion, infrequent cleaning, and health: The Situation in Developing Countries. Ann. Rev. Energy overcharging (36)) further suggest that emitted particles Environ. 1993, 18, 529-566.

86 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 1, 2007 (7) USEPA. Health Assessment Document for Diesel Engine Exhaust; (26) Gustafsson, O.; Haghseta, F.; Chan, H.; Macfarlane, J.; Gschwend, EPA/600/8-90./057F/; National Center for Environmental As- P. M. Quantification of the dilute sedimentary soot phase: sessment: Washington, DC, 2002. Implications for PAH speciation and bioavailability. Environ. (8) Bellouin, N.; Boucher, O.; Haywood, J.; Reddy, M. S. Global Sci. Technol. 1997, 31, 203-209. estimate of aerosol direct radiative forcing from satellite (27) Cooke, W. F.; Liousse, C.; Cachier, H.; Feichter, J. Construction measurements. Nature 2005, 438, 1138-1141. of a 1° 1° fossil fuel emission data set for carbonaceous aerosol (9) Streets, D. G.; Gupta, S.; Waldhoff, S. T.; Wang, M. Q.; Bond, T. and implementation× and radiative impact in the ECHAM4 C.; Yiyun, B. Black carbon emissions in . Atmos. Environ. model. J. Geophys. Res. 1999, 104, 22137-22162. 2001, 35, 4281-4296. (28) Chen, Y.; Sheng, G.; Bi, X.; Feng, Y.; Mai, B.; Fu, J. Emission (10) Novakov, T.; Hansen, J. E. Black carbon emissions in the United factors for carbonaceous particles and polycyclic aromatic Kingdom during the past four decades: an empirical analysis. hydrocarbons from residential coal combustion in China. Atmos. Environ. 2004, 38, 4155-4163. Environ. Sci. Technol. 2005, 39, 1861-1867. (11) Novakov, T.; Ramanathan, V.; Hansen, J. E.; Kirchstetter, T. W.; Sato, M.; Sinton, J. E.; Sathaye, J. A. Large historical changes of (29) Rogge, W. F.; Hildemann, L. M.; Mazurek, M. A.; Cass, G. R.; fossil-fuel black carbon aerosols. Geophys. Res. Lett. 2003, 30, Simoneit, B. R. T. Sources of Fine Organic Aerosol. 2. Noncatalyst 57. and Catalyst-Equipped Automobiles and Heavy-Duty Diesel (12) Chillrud, S. N.; Bopp, R. F.; Simpson, H. J.; Ross, J. M.; Shuster, Trucks. Environ. Sci. Technol. 1993, 27, 636-651. E. L.; Chaky, D. A.; Walsh, D. C.; Choy, C. C.; Tolley, L. R.; Yarme, (30) Lima, A. L. C.; Eglinton, T. I.; Reddy, C. M. High-resolution record A. Twentieth century atmospheric metal fluxes into Central Park of pyrogenic polycyclic aromatic hydrocarbon deposition during Lake, New York City. Environ. Sci. Technol. 1999, 33, 657-662. the 20th century. Environ. Sci. Technol. 2003, 37, 53-61. (13) Van Metre, P. C.; Mahler, B. J.; Furlong, E. T. Urban Sprawl (31) Hites, R. A.; Laflamme, R. E.; Windsor, J. G.; Farrington, J. W.; Leaves Its PAH Signature. Environ. Sci. Technol. 2000, 34, 4064- Deuser, W. G. Polycyclic Aromatic-Hydrocarbons in an Anoxic 4070. Sediment Core from the Pettaquamscutt River (Rhode-Island, (14) Routh, J.; Meyers, P. A.; Gustafsson, O.; Baskaran, M.; Hallberg, USA). Geochim. Cosmochim. Acta 1980, 44, 873-878. R.; Scholdstrom, A. Sedimentary geochemical record of human- induced environmental changes in the Lake Brunnsviken (32) Kroenke, A. E. Atmospheric mercury deposition to sediments watershed, Sweden. Limnol. Oceanogr. 2004, 49, 1560-1569. of New Jersey and Southern New York State : interpretations (15) Wakeham, S. G.; Forrest, J.; Masiello, C. A.; Gelinas, Y.; Alexander, from dated sediment cores. Ph.D. Dissertation, Rensselaer C. R.; Leavitt, P. R. Hydrocarbons in Lake Washington Sediments. Polytechnic Institute: Troy, NY, 2003; p 283. A 25-Year Retrospective in an Urban Lake. Environ. Sci. Technol. (33) National Research Council. Waste Incineration and Public 2005, 38, 431-439. Health; National Academy Press: Washington DC, 2000. (16) Yan, B. Z.; Abrajano, T. A.; Bopp, R. F.; Chaky, D. A.; Benedict, (34) Jonker, M. T. O.; Koelmans, A. A. Sorption of polycyclic aromatic L. A.; Chillrud, S. N. Molecular Tracers of Saturated and Polycyclic hydrocarbons and polychlorinated biphenyls to soot and soot- Aromatic Hydrocarbon Inputs into Central Park Lake, New York like materials in the aqueous environment. Mechanistic con- City. Environ. Sci. Technol. 2005, 39, 7012-7019. siderations. Environ. Sci. Technol. 2002, 36, 3725-3734. (17) Chaky, D. A. Polychlorinated biphenyls, polychlorinated dibenzo- (35) Walsh, D. C. The evolution of refuse incineration. Environ. Sci. p-dioxins and furans in the New York Metropolitan area: Technol. 2002, 36, 316A-322A. interpreting atmospheric deposition and sediment chronologies. Ph.D. Dissertation, Rensselaer Polytechnic Institute: Troy, NY, (36) Walsh, D. C.; Chillrud, S. N.; Simpson, H. J.; Bopp, R. F. Refuse 2003; p 431. incinerator particulate emissions and combustion residues for (18) Hedges, J. I.; Stern, J. H. Carbon and nitrogen determination of New York City during the 20th century. Environ. Sci. Technol. carbonate containing solids. Limnol. Oceanogr. 1984, 29, 657- 2001, 35, 2441-2447. 663. (37) Meyers, P. A. Applications of organic geochemistry to (19) Schmidt, M. W. I.; Skjemstad, J. O.; Czimczik, C. I.; Glaser, B.; paleolimnological reconstructions: a summary of examples Prentice, K. M.; Gelinas, Y.; Kuhlbusch, T. A. J. Comparative from the Laurentian Great Lakes. Org. Geochem. 2003, 34, 21- analysis of black carbon in soils. Global Biogeochem. Cycles 2001, 289. 15, 163-167. (38) Teranes, J. L.; Bernasconi, S. M. The record of nitrate utilization (20) Currie, L. A.; Benner, B. A. J.; Kessler, J. D.; Klinedinst, D. B.; and productivity limitation provided by delta N-15 values in Klouda, G. A.; Marolf, J. V.; Slater, J. F.; Wiseman, S. A.; Cachier, lake organic matter - A study of sediment trap and core H.; R., C.; Chow, J. C.; Watson, J.; Druffel, E. R. M.; Masiello, C. sediments from Baldeggersee, Switzerland. Limnol. Oceanogr. A.; Eglinton, T. I.; Pearson, A.; Reddy, C. M.; Gustafsson, O.; 2000, 45, 801-813. Quinn, J. G.; Hartmann, P. C.; Hedges, J. I.; Prentice, I. C.; (39) Jeong, G. Y.; Kim, S. J.; Chang, S. J. Black carbon pollution of Kirchstetter, T. W.; Novakov, T.; Puxbaum, H.; Schmid, H. A speleothems by fine urban aerosols in tourist caves.Am. Mineral. critical evaluation of interlaboratory data on total, elemental, 2003, 88, 1872-1878. and isotopic carbon in the carbonaceous particle reference material, NIST SRM 1649a. J. Res. Natl. Inst. Stand. Technol. (40) Ferrari, S.; Belevi, H.; Baccini, P. Chemical speciation of carbon 2002, 107, 279-298. in municipal solid waste incinerator residues. Waste Manage. (21) Gustafsson, O.; Bucheli, T. D.; Kukulska, Z.; Andersson, M.; 2002, 22, 303-314. Largeau, C.; Rouzaud, J.-N.; Reddy, K. R.; Eglington, T. I. (41) Jay, K.; Stieglitz, L. Identification and Quantification of Volatile Evaluation of a protocol for the quantification of black carbon Organic-Components in Emissions of Waste Incineration Plants. in sediments. Global Biogeochem. Cycles 2001, 15, 881-890. Chemosphere 1995, 30, 1249-1260. (22) Gelinas, Y.; Prentice, K. M.; Baldock, J. A.; Hedges, J. I. An (42) Zhou, H.-C.; Zhong, Z.-P.; Jin, B.-S.; Huang, Y. J.; Xiao, R. Improved Thermal Oxidation Method for the Quantification of Experimental study on the removal of PAHs using in-duct Soot/Graphitic Black Carbon in Sediments and Soils. Environ. activated carbon injection. Chemosphere 2005, 59, 861-869. Sci. Technol. 2001, 35, 3519-3525. (23) Nguyen, T. H.; Brown, R. A.; Ball, W. P. An evaluation of thermal (43) Knicker, H.; Gonzalez-Vila, F. J.; Polvillo, O.; Gonzalez, J. A.; resistance as a measure of black carbon content in diesel soot, Almendros, G. Fire-induced transformation of C- and N-forms wood char, and sediment. Org. Geochem. 2004, 217-234. in different organic soil fractions from a Dystric Cambisol under (24) Mitra, S.; Bianchi, T. S.; McKee, B.; Sutula, M. Black Carbon a Mediterranean pine forest (Pinus pinaster). Soil Biol. Biochem. - from the Mississippi River: Quantities, Sources, and Potential 2005, 37, 701 718. Implications for the Global Carbon Cycle. Environ. Sci. Technol. 2002, 36, 2296-2302. Received for review May 31, 2006. Revised manuscript re- (25) Eisenbud, M. Levels of exposure to sulfur oxides and particulates ceived September 12, 2006. Accepted October 10, 2006. in New York City and their sources. N.Y. Acad. Med. 1978, 54, 991-1011. ES061304+

VOL. 41, NO. 1, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 87