Environ. Sci. Technol. 2007, 41, 3864-3870
Fingerprinting Localized Dioxin Contamination: Ichihara Anchorage Case MINORI UCHIMIYA,* MARI ARAI, AND SHIGEKI MASUNAGA Graduate School of Environment and Information Sciences, Yokohama National University, 79-7 Tokiwadai, Hodogaya, Yokohama 240-8501, Japan
Although concentrations of polychlorinated dibenzo-pdioxins and dibezofurans (PCDD/Fs; dioxins) in majority of Japanese river and ocean sediments decreased below the national environmental quality standard of 150 pg-TEQ‚ (g-dry sediment)-1 by 2004, localized contamination inasmuch as 100-fold excess of the environmental quality standard has been reported at various locations including Ichihara Anchorage in northeastern Tokyo Bay. In the present study, we analyzed all mono- to octachlorinated dioxins in 12 surface sediments from Ichihara Anchorage and applied positive matrix factorization (PMF) to quantitatively fingerprint the congener pattern and geographical distribution of a factor causing the localized contamination. A PMFderived fingerprint attributable to dioxin impurities in pentachlorophenol (PCP) exerted more than 90% contribution to total dioxin concentrations in Ichihara Anchorage surface sediments. Although majority of Ichihara Anchorageborn dioxins were trapped at the origin, contribution of the PCP-derived dioxins in overall Tokyo Bay gradually increased toward Ichihara Anchorage, indicating the impact of localized dioxin contamination on a large proportion of Tokyo Bay. We suggest that, in addition to runoff from rice paddies (to which PCP had long been applied as herbicide) at the basin, Ichihara Anchorage serves as a significant source of PCP-derived dioxins especially in eastern Tokyo Bay.
Introduction After the national environmental quality standard of dioxin concentrations in toxic equivalents (TEQ) was set to 150 pg-TEQ‚(g-dry sediment)-1 in 2000, concentrations of polychlorinated dibenzo-p-dioxins and dibezofurans (PCDD/Fs; dioxins) in Japanese river and ocean sediments has been on a steady fall down to 7.5 pg-TEQ‚(g-dry sediment)-1 (national average) by 2004 (1). On the other hand, localized dioxin contamination inasmuch as 100-fold excess of the environmental quality standard has been reported at various locations including Ichihara Anchorage in northeastern Tokyo Bay (Figure 1) (1). As an initial step toward the remediation of these highly contaminated sites, regional governments are calling for the development of quantitative field monitoring tools to first determine the specific causes of the localized contaminations. Tokyo Bay is an inner bay in the Tokyo Metropolitan area surrounded by various industrial plants such as oil refineries * Corresponding author phone: +81-45-339-4346; fax: +81-45339-4352; e-mail:
[email protected]. 3864
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 11, 2007
and power plants, municipal solid waste incineration facilities, as well as agricultural fields that comprise a considerable percentage of the catchment area (approximately 20%) (2). Within the areas of Tokyo Bay in which the environmental quality standard is met, comprehensive studies have been carried out to quantify all tetra- to octachlorinated PCDD/Fs in surface sediments (3) and a core (4) as well as all monoto octachlorinated PCDD/Fs in aqueous dissolved and particulate phases (5). For PCDD/Fs in aqueous particulate and dissolved phases within Tokyo Bay, a dynamic box model called FATE3D was employed to predict the seasonal trends in TEQ (6) by considering the partitioning between dissolved and particulate phases, sedimentation of the particulate phase, advection, and diffusion (6). Two input variables, river inflows and atmospheric deposition, adequately accounted for total dissolved and particulate PCDD/F concentrations within the bay (6). In support of this modeling effort, principal component analysis (PCA) and chemical mass balance method (CMB) on PCDD/Fs in aqueous dissolved and particulate phases (5) as well as core (4) and surface sediments (7) suggested the following primary sources of PCDD/Fs in Tokyo Bay: combustion and PCDD/F-contaminated herbicides pentachlorophenol (PCP) and chloronitrophen (CNP) widely applied on rice paddies in Japan between late 1950s and early 1990s (8). These results suggest that, apart from combustion, PCP- and CNP-derived PCDD/Fs in runoff from rice paddies in the basin constitute majority of the contamination scenarios in parts of Tokyo Bay having TEQ below the environmental quality standard. At Ichihara Anchorage, in contrast, only PCDD/Fs fully chlorinated at 2,3,7,8-positions have been quantified (to determine TEQ in Figure 1), and to our knowledge, no quantitative evidence exists on the congener profiles and geographical distributions of source(s) responsible for the localized contamination. The objective of this study is to quantitatively determine the factors causing the localized dioxin contamination in Ichihara Anchorage in northeastern Tokyo Bay (Figure 1) and to evaluate its impact on the entire Tokyo Bay. Receptor models such as the positive matrix factorization (PMF) are powerful statistical tools for quantitatively resolving the number, chemical composition, and geographical/temporal distribution of the chemical fingerprints simultaneously (9). Receptor models have been widely employed in air pollution studies (e.g., refs 10, 11), and have recently found significant application to investigate sediment pollution (12-14). Our preceding report in this series employed PMF to determine the time trends in sources and dechlorination pathways of PCDD/Fs in sediment cores from Tokyo Bay and Lake Shinji (14). In the present study, we performed a rigorous quantification of all mono- to octachlorinated dioxins in surface sediments of Ichihara Anchorage and employed PMF to quantitatively fingerprint the chemical compositions and geographical distributions of factors causing the localized contamination. Our two reports in this series revealed integral roles played by the dynamics, spatiotemporal distribution of PMF-derived fingerprints in environmental forensics.
Materials and Methods Sample Collection and Analysis. Surface sediments were collected at five locations in Ichihara Anchorage and seven locations in the immediate surrounding areas in April 2002. To prevent sample contamination/aging, sediment samples were immediately freeze-dried and stored in the dark at -30 °C until analyses. Methods employed for the collection of 10.1021/es062998p CCC: $37.00
2007 American Chemical Society Published on Web 04/21/2007
FIGURE 1. Localized dioxin contamination in sediments (in pg-TEQ‚(g-dry sediment)-1; brown) and surface water (in pg-TEQ‚L-1; blue) of Japan. Values represent the maximum concentrations detected at each location.
TABLE 1. Homologue-specific Concentrations of PCDD/Fs and co-PCBs in 12 Sites (I1-I12) Surrounding Ichihara Anchorage of Northeastern Tokyo Bay A. Total Mono- to Octachlorinated PCDD/Fs in pg‚(g-dry sediment)-1 I3 I4 I5 I6 I7 I8
homologue
I1
I2
monoCDFs diCDFs triCDFs TeCDFs PeCDFs HxCDFs HpCDFs OcCDF monoCDDs diCDDs triCDDs TeCDDs PeCDDs HxCDDs HpCDDs OcCDD total
10,612 27,108 82,820 20,740 22,824 249,086 1,173,906 1,116,102 89 831 654 6,227 3,969 40,309 1,257,323 8,132,109 12,144,708
399 779 3,014 4,795 5,209 44,151 97,940 143,928 52 647 433 25,597 1,969 8,535 121,707 825,724 1,284,879
262 185 497 834 989 16,527 44,515 63,421 43 669 161 3,757 688 3,490 55,737 361,275 553,051
I12
24 9 15 21 21 65 142 139 7 79 19 509 57 65 333 1,892 3,396
49 41 79 184 146 1,289 2,956 3,562 21 213 110 4,096 596 335 4,738 35,387 53,803
87 100 120 440 283 627 1,536 1,676 60 573 501 10,450 2,135 374 2,442 14,370 35,776
61 62 79 150 156 589 1,366 1,714 39 353 119 1,798 492 234 2,182 14,959 24,351
B. Total Co-PCBs in pg‚(g-dry sediment)-1 I5 I6 I7 I8
I9
I10
I11
I12
374
11,918
4,833
3,402
I3
I4
78,497
16,087
12,860
24,341
5,506
3,537
89 182 627 2,710 1,586 11,619 22,025 31,296 36 234 181 2,805 437 2,079 22,721 161,358 259,986
I11
42 32 83 169 124 840 2,041 2,714 31 218 107 1,975 239 234 2,750 19,550 31,151
I2
104 81 157 295 338 3,293 10,049 13,579 52 554 178 3,597 627 981 12,320 88,198 134,401
I10
105 93 183 373 283 1,790 4,216 5,484 73 514 294 7,174 1,036 487 6,006 45,806 73,917
I1
263 203 425 790 947 12,911 40,970 64,206 66 682 232 5,259 966 3,177 43,157 311,929 486,181
I9
5,564
1,399 sediment)-1
C. Total PCDD/Fs and Co-PCBs in pg-TEQ‚(g-dry I3 I4 I5 I6 I7
homologue
I1
I2
total PCDD/Fs total co-PCBs total
15,472 14.8 15,487
1,670 3.6 1,674
750 3.4 753
603 7.1 610
sediments, cleanup procedures, and HRGC/HRMS analyses of tetra- to octachlorinated PCDD/Fs and coplanar poly-
172 1.9 174
331 1.0 332
90 1.9 92
I8
I9
I10
I11
I12
40 0.5 40
9 0.2 9
66 3.4 70
53 2.2 55
43 1.5 44
chlorinated biphenyls (co-PCBs) are described in a previous report (15). Mono- to trichlorinated dioxins were quantified VOL. 41, NO. 11, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
3865
FIGURE 2. PCDD/F and co-PCB concentrations (in pg-TEQ‚(g-dry sediment)-1) in Ichihara Anchorage (I1-12; present study) and in entire Tokyo Bay (S1-7; ref 3). All bars are to scale with one another. using a SP-2331 column. A total of 149 chromatographic peaks for mono- to octachlorinated PCDD/Fs and 12 co-PCBs were quantified in each sample. Mean recoveries were 60-100% ((4-19%) for tetra- to octachlorinated PCDD/Fs fully chlorinated at 2,3,7,8-positions and co-PCBs (except 37 ( 17% for no. 114). For mono- to trichlorinated dioxins, mean recoveries were 20 ( 12% for 2-MonoCDF, 46 ( 13% for 2,8-DiCDF, 64 ( 9% for 2,4,8-TriCDF, 23 ( 12% for 2-MonoCDD, 44 ( 13% for 2,3-DiCDD, and 68 ( 10% for 2,3,7-TriCDD. Only trace amounts of PCDD/Fs were detected in method blanks. Positive Matrix Factorization. The PMF is a PCA-based receptor model with nonnegativity constraints that involve solution of quantitative source apportionment equations by the oblique solutions in reduced dimensional space (16). The following linear algebraic equation addresses PMF (17): p
xij )
∑a
ikfkj
+ ij
(1)
k)1
where xij is the concentration of the jth congener in ith sample of the original data set, aik is the contribution of the kth 3866
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 11, 2007
factor on sample i, fkj is the fraction of the kth factor arising from congener j, and ij is the residual between xij and the estimate of xij using p principal components. The objective of PMF is to minimize Q, the weighted sum of squares of differences between the PMF output and the original data set (17):
( )
2
p
n
Q)
m
∑∑ i)1 j)1
xij -
∑a k)1
sij
ikfkj
(2)
where sij is the uncertainty of the jth congener in ith sample of the original data set containing m congeners and n samples. A computer software program (18) based on Paatero’s PMF program (19) was used in robust mode. For more exact replications of the simulated data, predicted factor contributions were allowed to be slightly negative down to -0.1 (18). For each run, 30 random starting points afforded comparable Q values, indicating that the numerical solutions were found at the global, rather than the local, minima (18). Pretreatment of Data Sets. Original data set from the present study (12 Ichihara Anchorage surface sediments) was
FIGURE 3. Congener patterns of mono- to octachlorinated PCDD/Fs (in ng/g-dry sediment) in Ichihara Anchorage surface sediments exceeding the environmental quality standard (I1-6 in Figure 2). Numbers represent different congeners. Vertical lines separate PCDDs and PCDFs. first combined with literature values containing seven surface sediments collected from a broader region of Tokyo Bay (3). Combined data set (in pg‚(g-dry sediment)-1) was converted to % contribution to total tetra- to octachlorinated PCDD/F concentrations in each sample to normalize out the effects of dilution away from the source, thereby allowing precise recognition of the relative proportions of congeners in the fingerprint. Then, values below the detection limit (DL) were replaced with half of DL (9). Congeners with more than 15% below DL values were eliminated. The final data set contained 19 surface sediment samples and 83 chromatographic peaks for tetra- to octachlorinated PCDD/Fs. Co-PCBs were not included in the PMF analysis. Because the PMF analyses involved literature values and data sets were normalized to the total PCDD/F concentration of each sample, the uncertainty of the jth congener in ith sample (sij value in eq 2) was fixed to the standard deviation of the jth congener for 19 samples considered for the PMF analysis. The PMF analysis of normalized data set afforded the congener profile (fkj in eq 1) and contribution (aik) for each
factor. For a given factor, the aikfkj values (% factor contribution to total PCDD/F concentrations) allow us to recognize samples having significant contributions from a fingerprint regardless of total PCDD/F concentrations. Absolute factor contributions (in pg‚(g-dry sediment)-1) can then be determined using total PCDD/F concentrations. Sum of contributions for all factors canceled out negative contributions to better reproduce the original data set. Error Analyses. The coefficient of determination (COD) was used to evaluate the ability of PMF to reproduce the original data set, and to determine the number of principal components. The COD provides the goodness of fit (r2) between the observed and predicted concentration of each congener and equals 1.0 for a perfect fit (9).
Results and Discussion Dioxins in Ichihara Anchorage. Figure 2 presents the geographical distribution of TEQ arising from PCDD/Fs and co-PCBs (in pg-TEQ‚(g-dry sediment)-1) in Ichihara Anchorage (I1-12). Figure 2 also shows previously reported TEQ (3) VOL. 41, NO. 11, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
3867
FIGURE 4. Chemical compositions of fingerprints obtained from a data set containing 12 surface sediments in Ichihara Anchorage and seven surface sediments from broader areas of Tokyo Bay. Vertical lines separate PCDDs and PCDFs. in additional surface sediments collected in 1995 (S1-7). While TEQ in the majority of Tokyo Bay (S1-7 and I7-12) are well below the national environmental quality standard of 150 pg-TEQ‚(g-dry sediment)-1, TEQ at sites I1-6 are as much as 100-fold greater. Labels in Figure 2 (S1-7 and I1-12) will be used to denote each sampling site throughout this paper. Homologue-specific concentrations of PCDD/Fs and PCBs at sites I1-12 are given in Table 1 (in both TEQ and absolute concentration). Table 1 clearly indicates localized contamination of both PCDD/Fs and PCBs in Ichihara Anchorage. Total concentration (in pg‚(g-dry sediment)-1) of mono- to octachlorinated dioxins peaked at the innermost Ichihara Anchorage (site I1) and sequentially dropped by several orders of magnitude toward outer anchorage, resulting inasmuch as 3600-fold greater concentration at site I1 compared to the immediately surrounding areas (Table 1A). Figure 3 presents congener patterns of mono- to octachlorinated dioxins (in ng/g-dry sediment) in Ichihara Anchorage surface sediments I1-6 exhibiting TEQ above the environmental quality standard. Congener patterns of surface sediments I1-6 closely resemble one another and are dominated by OcCDD, heptachlorinated PCDD/Fs (1,2,3,4,6,7,9HpCDD, 1,2,3,4,6,7,8-HpCDD, and 1,2,3,4,6,8,9-HpCDF) and OcCDF. This congener profile characterizes the PCDD/F byproducts formed during the chlorination of phenol to form PCP (Figure S1, Supporting Information) (8). It is hypothesized that OcCDD is formed as a coupling product of two PCP molecules (8). Heptachlorinated homologues can be formed via the coupling between PCP and the major chlorination intermediate 2,3,4,6-tetrachlorophenol (8). The OcCDF is likely formed by the coupling of two PCP molecules 3868
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 11, 2007
(8). The 1,2,4,6,8,9-HxCDF, a minor HxCDF contributor in I2, I3, I5, and I6 (Figure 3), can be formed by the coupling of two 2,3,4,6-tetrachlorophenol intermediates (8). Indeed, congener patterns of surface sediments I1-6 are nearly indistinguishable from the reference PCP congener profile (Figure S1A) not only with respect to heptachlorinated PCDD/ Fs and OcCDF but to % contribution of OcCDD to total PCDD/F concentrations (65 ( 2% in I1-6 and 60 ( 28% in Figure S1A). Possible sources of 1,3,6,8-TeCDD, a minor contributor in I2 and I5 (Figure 3), are discussed in the following sections. Congener patterns for sites I7-12 are provided in Figure S2 of the Supporting Information Section. PMF Analysis. To quantitatively resolve the chemical composition and distribution of sources causing the localized dioxin contamination in Ichihara Anchorage, PMF analysis was performed on a combined data set from our present study (I1-12 in Figure 2) and literature values (S1-7 in Figure 2) (3). Because the literature values only contained concentrations of tetra- to octachlorinated PCDD/Fs (3) and only trace amounts of mono- to trichlorinated dioxins were observed in Ichihara Anchorage surface sediments (Figures 3 and S2), mono- to trichlorinated dioxins were not included in the PMF analysis. The PMF analysis of the combined data set (83 chromatographic peaks of tetra- to octachlorinated PCDD/Fs for 19 surface sediments) afforded four fingerprints. As described in a previous receptor model-based study (12), disproportionately high concentrations of OcCDD in sediments (61 ( 6% contribution to total PCDD/F concentrations for 19 samples included in our PMF analysis) cause its dominant, artifact presence in most factors. In the present study, identities of PMF-derived factors were deduced using literature source profiles by taking into account all congeners except OcCDD. Previous studies suggested photochemical
FIGURE 5. The % factor contribution to total PCDD/F concentrations for CNP, PCP, and combustion fingerprints in Ichihara Anchorage (inlet) and broader areas of Tokyo Bay. All stacked bars are to scale with one another and sum to 100% at each site. Underlined values show total tetra- to octachlorinated PCDD/F concentrations in ng‚(g-dry sediment)-1. synthesis of OcCDD from PCP in rain (20) and in situ biotic/ abiotic formation of OcCDD from chlorophenol precursors (21) as the sources of excess OcCDD in sediments. In addition, analyses for mono- to octachlorinated dioxins in aqueous dissolved and particulate phases in river estuary (22) suggested the importance of partitioning between dissolved and particulate phases on the homologue distributions. Figure 4 shows the chemical compositions of four PMFderived fingerprints (fkj in eq 1). Figure 4A is dominated by the signature TeCDD byproducts of CNP synthesis from 2,4,6trichlorophenol and 4-chloronitrobenzene: 1,3,6,8- and 1,3,7,9-TeCDDs (8) (Figure S1B). The primary, 1,3,6,8substituted TeCDD is likely formed as a coupling product of two 2,4,6-trichlorophenols (8). The 1,3,7,9-TeCDD can be formed by the coupling of two 2,4,6-trichlorophenols followed by the Smiles rearrangement (8). The second fingerprint (Figure 4B) closely resembled the congener profiles of PCDD/ Fs arising from waste incineration processes in Kanto Region of Japan (23) (Figure S1C) and was attributed to combustion. In addition, two fingerprints having nearly identical congener patterns representative of PCP were obtained (Figure 4C-D). Congener patterns of the two PCP fingerprints do not sufficiently differ form one another to discuss possible
origins (e.g., different synthetic pathways of PCP leading to different PCDD/F byproducts) and their contributions are combined to represent one PCP fingerprint. By increasing the number of factors from three to four, COD (Table S1, Supporting Information) of all congeners except 1,2,3,9-TeCDF and 1,2,3,6,7,8-HxCDF improved to 0.6 and above. The fifth factor afforded an additional PCP fingerprint and was not considered. When the number of factors was reduced to three, COD of the main congeners for CNP, 1,3,6,8-TeCDD and 1,3,7,9-TeCDD, significantly decreased (Table S1) to afford two fingerprints that are a mix of CNP and combustion fingerprints, in addition to a PCP fingerprint. Figure 5 provides % factor contribution to total PCDD/F concentrations (stacked bars sum to 100% at each site) for CNP, PCP, and combustion fingerprints (Figure 4). The PCP fingerprint exerts exclusive (more than 90%) contributions on sites exceeding the environmental quality standard (I1-6). Correspondingly, PCDD/F congener patterns of sites I1-6 (Figure 3) closely resemble that of PCP reference (Figure S1A), indicating minimal influence of CNP and combustion at these sites. On the other hand, combustion and CNP fingerprints contribute significantly to sites that meet the VOL. 41, NO. 11, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
3869
environmental quality standard (S1-7 and I7-12 in Figure 5). Correspondingly, congener patterns of sites I7-12 show strong contributions from additional congeners, especially 1,3,6,8-TeCDD and 1,3,7,9-TeCDD (Figure S2). In particular, the CNP fingerprint (Figure 4A) is nearly indistinguishable from the congener profile of sample I11 (Figure S2) that receives exclusive contribution from the CNP fingerprint (Figure 5). Presence of heptachlorinated PCDD/Fs and OcCDF in both the CNP fingerprint (Figure 4A) and sample I11 (Figure S2), but not in the reference (Figure S1B), indicates different synthetic pathways of CNP leading to different PCDD/F impurities and/or the impact of PCP. Total tetra to octachlorinated PCDD/F concentrations (in pg‚(g-dry sediment)-1) attributable to the PCP fingerprint are at least 3 orders of magnitude greater at sites I1-6 compared to S1-7 and I7-12 (Figure S3, Supporting Information), suggesting that majority of Ichihara Anchorage-born PCDD/Fs stays trapped at the origin. A dynamic box-model-based simulation suggested that river inflows (carrying runoff from rice paddies in the basin) are the major sources of PCP-derived PCDD/Fs in the particulate phase (which undergoes sedimentation) in western Tokyo Bay (6). In the present study, % factor contribution of the PCP fingerprint at site S1-7 increased toward Ichihara Anchorage (Figure 5). While our PMF analysis did not distinguish PCP-derived PCDD/Fs from two putative origins (river inflows and Ichihara Anchorage), higher contribution of PCP-derived PCDD/Fs toward Ichihara Anchorage (S1-7 in Figure 5) indicated an impact of localized contamination on broader areas of Tokyo Bay. Previous CMB analysis on surface sediments S1-7 also suggested greater contribution of PCP toward eastern Tokyo Bay (7). Although the agricultural use of PCP was banned in 1990, 164 ktons (national sum) of dioxin-contaminated PCP used in rice paddies between late 1950s and late 1980s (8) left the basin a terrestrial reservoir of dioxins. Our present study provided an evidence for a separate, localized reservoir of PCP-derived dioxins in Ichihara Anchorage of northeastern Tokyo Bay. Ichihara Anchorage was developed in 1959 as a part of the reclamation effort to construct the Chiba Prot (see map in Figure 5). The operation of chemical (in addition to petroleum refining, steel, and electric) factories on Ichihara Anchorage began in 1960. Factory effluent from a PCPmanufacturing plant likely created the local well of PCPderived dioxins, though the possibility of solid waste/sludge disposal cannot be ruled out.
Acknowledgments This work was supported by the 21st Century COE Program “Environmental Risk Management for Bio/Eco-systems” of Ministry of Education, Culture, Sports, Science and Technology of Japan.
(4)
(5) (6)
(7) (8) (9)
(10) (11)
(12)
(13)
(14) (15)
(16) (17) (18) (19) (20)
(21)
Supporting Information Available Diagnostic tools for the PMF analysis, reference congener profiles of PCP, CNP, and combustion, congener profiles of surface sediments I7-12, and total PCDD/F concentrations arising from CNP, PCP, and combustion fingerprints. This material is available free of charge via the Internet at http:// pubs.acs.org.
Literature Cited (1) Environmental Agency of Japan. Dioxins Emission Inventory; , 2000-2004 (in Japanese). (2) National Land Agency. Development of Metropolitan Areas; 1995 (in Japanese). (3) Sakurai, T.; Kim, J. G.; Suzuki, N.; Matsuo, T.; Li, D. Q.; Yao, Y. A.; Masunaga, S.; Nakanishi, J. Polychlorinated dibenzo-p-
3870
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 11, 2007
(22)
(23)
dioxins and dibenzofurans in sediment, soil, fish, shellfish and crab samples from Tokyo Bay area, Japan. Chemosphere 2000, 40, 627-640. Masunaga, S.; Yao, Y.; Ogura, I.; Sakurai, T.; Nakanishi, J. Source and behavior analyses of dioxins based on congener-specific information and their application to Tokyo Bay basin. Chemosphere 2003, 53, 315-324. Kobayashi, K.; Masunaga, S.; Nakanishi, J. Source identification of PCDD/Fs and dioxin-like PCBs in rivers flowing into the Tokyo Bay, Japan. Organohalogen Compd. 2003, 61, 361-364. Kobayashi, N.; Eriguchi, T.; Nakata, K.; Masunaga, S.; Horiguchi, F.; Nakanishi, J. Application of a 3-D chemical fate prediction model (FATE3D) to predict dioxin concentrations in the Tokyo Bay. Estuarine, Coastal Shelf Sci. 2006, 70, 621-632. Sakurai, T. Dioxins in aquatic sediment and soil in the Kanto region of Japan: major sources and their contributions. Environ. Sci. Technol. 2003, 37, 3133-3140. Masunaga, S.; Takasuga, T.; Nakanishi, J. Dioxin and dioxinlike PCB impurities in some Japanese agrochemical formulations. Chemosphere 2001, 44, 873-885. Johnson, G. W.; Ehrlich, R.; Full, W. Principal components analysis and receptor models in environmental forensics. In An Introduction to Environmental Forensics; Murphy, B. L., Morrison, R. D., Eds.; Academic Press: San Diego, 2002; pp 461515. Kim, E.; Hopke, P. K.; Larson, T. V.; Covert, D. S. Analysis of ambient particle size distributions using unmix and positive matrix factorization. Environ. Sci. Technol. 2004, 38, 202-209. Poirot, R. L.; Wishinski, P. R.; Hopke, P. K.; Polissar, A. V. Comparative application of multiple receptor methods to identify aerosol sources in northern Vermont. Environ. Sci. Technol. 2001. Barabas, N.; Goovaerts, P.; Adriaens, P. Modified polytopic vector analysis to identify and quantify a dioxin dechlorination signature in sediments. 2. application to the Passaic River. Environ. Sci. Technol. 2004, 38, 1821-1827. Bzdusek, P. A.; Lu, J. H.; Christensen, E. R. PCB congeners and dechlorination in sediments of Sheboygan River, Wisconsin, determined by matrix factorization. Environ. Sci. Technol. 2006, 40, 120-129. Uchimiya, M.; Masunaga, S. Time trends in sources and dechlorination pathways of dioxins in agrochemically contaminated sediments of Japan. Environ. Sci. Technol., in press. Yao, Y.; Masunaga, S.; Takada, H.; Nakanishi, J. Identification of polychlorinated dibenzo-p-dioxin, dibenzofuran, and coplanar polychlorinated biphenyl sources in Tokyo Bay, Japan. Environ. Toxicol. Chem. 2002, 21, 991-998. Johnson, G. W.; Ehrlich, R. State of the art report on multivariate chemometric methods in environmental forensics. Environ. Forensics 2002, 3, 59-79. Paatero, P.; Tapper, U. Positive matrix factorization: a nonnegative factor model with optimal utilization of error estimates of data values. Environmetrics 1994, 5, 111-126. EPA PMF 1.1 for Windows; U.S. EPA: Research Triangle Park, NC, 2005. Paatero, P. Least squares formulation of robust non-negative factor analysis. Chemom. Intell. Lab. Syst.1997, 37, 23-35. Baker, J. I.; Hites, R. A. Is combustion the major source of polychlorinated dibenzo-p-dioxins and dibenzofurans to the environment? a mass balance investigation. Environ. Sci. Technol. 2000, 34, 2879-2886. Gaus, C.; Brunskill, G. J.; Connell, D. W.; Prange, J.; Muller, J. F.; Papke, O.; Weber, R. Transformation processes, pathways, and possible sources of distinctive polychlorinated dibenzop-dioxin signatures in sink environments. Environ. Sci. Technol. 2002, 36, 3542-3549. Lohmann, R.; Nelson, E.; Eisenreich, S. J.; Jones, K. C. Evidence for dynamic air-water exchange of PCDD/Fs: a study in the Raritan Bay/Hudson River estuary. Environ. Sci. Technol. 2000, 34, 3086-3093. Ogura, I.; Masunaga, S.; Nakanishi, J. Congener-specific characterization of PCDDs/PCDFs in atmospheric deposition: comparison of profiles between deposition, source and environmental sink. Chemosphere 2001, 45, 173-183.
Received for review December 18, 2006. Revised manuscript received March 19, 2007. Accepted March 22, 2007. ES062998P
