Environ. Sci. Technol. 1991, 25,310-316
Rare Earth Elements in Sediments off Southern California: A New Anthropogenic Indicator Ilhan Olmez
Massachusetts Institute of Technology, Nuclear Reactor Laboratory, Cambridge, Massachusetts 02 139 Edward I?.Sholkovltr”
Woods Hole Oceanographic Institution, Woods Hole, Massachusetts 02543 Diane Hermann
Massachusetts Institute of Technology, Nuclear Reactor Laboratory, Cambridge, Massachusetts 02 139 Robert P. Eganhouse
Southern California Coastal Water Research Project, 646 West Pacific Coast Highway, Long Beach, California 90806 The rare earth elements (REE) composition of sediment cores from the San Pedro Shelf (60 m) and the Santa Barbara Basin (588 m) are contrasted. The Santa Barbara Basin core has relatively uniform REE concentrations throughout its 60-cm length and a REE composition similar to the crustal abundance. In contrast, the upper 20 cm of the 36-cm San Pedro Shelf core collected in 1981 is enriched in the light REE (La, Ce, Nd, Sm) but not the middle REE (Eu) or the heavy REE (Yb, Lu). These upper sediments and two effluent particulate samples obtained in 1979 have REE signatures markedly different from crustal material. This is a result of anthropogenic inputs beginning in the early 1960s from the Joint Water Pollution Control Plant wastewater outfall 6 km upcurrent from the San Pedro Shelf core. The sources of the light REE enrichment are petroleum-cracking catalysts and their products; these include bottom ash, fly ash, and wastewater from oil-fired power plants and oil refineries. Cracking catalysts, which are produced primarily from two REE minerals, bastnasite and monazite, are strongly enriched in the light REE. With their unique signature and source, the REE may be a new tracer for anthropogenic inputs in coastal environments. 1. Introduction Rare earth elements (REE) form a chemically “coherent” series of elements, which decrease in ionic radii from the lighest element, La, to the heaviest element, Lu. Because of this “lanthanide contraction”, the relative abundances of REE in different types of rock, sediment, and water have characteristic signatures, which are used to study aquatic and solid-phase geochemistry (1-3). The signature or pattern of the REE is usually described by normalizing the individual REE concentrations of a sample to those of the crustal abundance of the earth. An average REE composition of shale is used for this normalization ( 1 , 2 ) . As an example of signature characterizations, seawater is enriched in the heavy REE relative to shale while granites are enriched in the light REE (1-3). For this paper La, Ce, Nd, and Sm represent the light (LREE), Eu and T b the middle (MREE), and Er, Yb, and Lu the heavy (HREE) rare earth elements. In spite of their name, rare earth elements are more naturally abundant in sediments than many of the trace metals commonly studied by geochemists and environmental chemists. Shales have La, Ce, and Nd concentrations between 40 and 80 ppm while Yb and Lu range between 0.5 and 4 ppm (2). By way of comparison, average background concentrations of Ag, Cd, Cr, Cu, Ni, Pb, and 310
Environ. Sci. Technol., Vol. 25, No. 2, 1991
Table I. Concentration (ppm) and Shale-Normalized Ratio (SNR) of Monazite and Bastnasite in Ore Deposits’ monazite PPm SNR La Ce Nd Sm Eu Gd DY Er Yb Lu
202000 389000 147000 22000 400 13000 5900 1800 1100 350
4930 4690 3870 2930 250 2050 1070 480 310 580
bastnasite PPm SNR 283000 400000 103000 6700 950 1500 270 31 11 0.9
6900 4820 2710 890 590 240 49 8.3 3.1 1.4
“Data from Table 12.1 of Neary and Highley (30). Monzanite from western Australia and Bastnasite from Mountain Pass, CA.
Zn in nonpolluted sediments from southern California marine sediments are 0.2, 0.4, 25, 9, 15, 10, and 44 ppm, respectively ( 4 ) . Improvements in the application of fluid-cracking catalysts in the petroleum industry resulted from incorporating REE into zeolite catalysts (5, 6). These catalysts were invented in the late 1950s, and the demand for REE-containing zeolites by the petroleum industry increased steadily in the early 1960s. These cracking catalysts are produced from two major REE ore minerals, bastnasite and monazite. The Mountain Pass (CA) bastnasite deposit is the major source of REE in the United States and 50% of the world’s production (7-9). As shown in Table I, these minerals are greatly enriched in the LREE (7-10). For bastnasite the concentrations of La, Ce, Nd, and Sm are 1000-6000 times greater than those of shale while the HREE-Er, Yb, and Lu-are enriched less than %fold. The compositions in Table I should not be taken as “the” REE abundances or ores used to produce oil-cracking catalysts. Because of the complexity and diversity of the mineralogy of the ore minerals and the proprietary nature of the zeolite manufacturing processes, Table I only illustrates that the REE signature of oilcracking catalysts will be distinctly different from that of the crust. Olmez and Gordon (11) have shown that atmospheric particles from oil-fired (but not coal) plants and refineries have LREE-enriched patterns due to zeolite cracking catalysts. Fly ash, bottom ash, and waters (ground and leachates) from oil-fired power plants also contain LREE-enriched signatures (12, 13). This paper is the first to report an anthopogenic REE signal of oil-refined petroleum products in aquatic sedi-
0013-936X/91/0925-0310$02.50/0
0 1991 American Chemical Society
31
Fig Location maps. (a) Submarine basins offsouthem Catiirnia including site of Sam Barbara Basin cure. (b) Location of San Pedro Sheif cure (3C1) and JWPCP outfall system.
ments. We show that there is a dramatic change in both the relative abundance and concentration of REE in a sediment profile from the shelf of the San Pedro Basin off southern California. The distinct REE signature in the upper 20 cm is derived from oil-refining cracking catalysts, their products, and oil-fired power plant emissions and their waste products (i.e., fly and bottom ash and leachates). Since we cannot identify specific phases in the sediments, we will use the broad term "refined petroleum products" to represent the many and related sources described above. The noncrustal signature of REE may provide a new tool for determining the deposition and transport of a source-specific contaminant in coastal regions. 2. Southern California Borderland Sediments of the "inner" borderland basins off southern California (Figure 1)-San Pedro, Santa Monica, and Santa Barbara Basins-have provided stratigraphic records of the anthropogenic input of trace organic and inorganic constituents (4,14-20). Being situated close to the densely populated and highly industrial region of Los Angeles, the Santa Monica and San Pedro Basins receive large amounts of anthropogenic inputs from wastewater effluent, storm runoff, and atmospheric deposition. Treated wastewater is discharged to the shelf of the San Pedro Basin from the Joint Water Pollution Control Plant (JWPCP) outfalls located 3 km offshore at depth of 60 m (14,18,20) (Figure lh). T o compare an outfall-impacted
sediment with a more distant site, we analyzed a core from the San Pedro Shelf and a core from the center of the Santa Barbara Basin (Figure 1). The shelf core (3C1) was collected in October 1981, and we analyzed the same sediments that Eganhouse and Kaplan (14) used to reconstruct the depositional history of organic carbon and several trace organic contaminants. Sediment horizons from the Santa Barbara Basin were provided by Dr. C. Reimers (Scripps Institution of Oceanography). This core came from the deep part of the hasin (588 m; 34'14'06'' N and 120'02'34" W) where anoxic conditions in the sediments result in undisturbed laminated sediments (21). Although sedimentation rates on this specific core were not determined, numerous studies in this basin yield rates of approximately 0.5 cmfyear for the upper meter (22). The shelf sampling site is located 6 km north of the JWPCP outfalls at the same water depth (60 m). The 1980 effluent from this outfall system was approximately 1.4 X lo9 L/day (14, 18,20). According to Stull et al., (23), this treatment plant processes domestic waste from approximately 3.5 million people and waste from tens of thousands of businesses and industries. In the late 1970s and early 1980s approximately 17 oil refineries were connected to the JWPCP wastewater system (W. Micklish, Los Angeles Sanitation District, personal communication). Sediments on the San Pedro Shelf surrounding and including our 3C1 core have been extensively studied with respect to their geological, biological, and chemical features. The transport of water and particles to the northwest along this shelf has resulted in deposition of JWPCP-derived inorganic and organic contaminants along the 60-m isobath. The core 3C1 profiles of organic carbon, organic nitrogen, and their stable isotopic composition unequivocally demonstrate that 80-90% of the organic matter in the upper 20 cm is derived from sewage effluent (14). These same sediments also contain greatly elevated concentrations of lead and other trace metals (18). Galloway's (20) seminal paper showed that only a small fraction (less than 10%) of wastewater-injected trace metals (e.g., Cd, Cr, Cu, Pb, Ag, and Zn) are retained near the outfall region. The large-scale transport of solids away from the discharge sites and deposition in basin sediments has been confirmed (4, 14, 16). Surface sediments from the center of the Santa Barbara Basin are much less impacted by anthropogenic trace metals and organic matter (15, 16). Not only are the southern sources of contaminants more distant, but this hasin also receives sediment from the north (241, a region of much less industry and people. 3. Samples and Analyses Core 3C1 had 1-2 cm of brown sediment overlying black sulfidic sediment. From a time series study of the San Pedro Shelf sediments in the 1970s, the upper sediments from our site varied from having no smell of hydrogen sulfide to having a moderate odor (23). We have no additional data with which to quantify the diagenetic (i.e., redox) and sedimentological (i.e., bioturbation rates) features of our core. The dates on Figure 2 for core 3C1 are primarily derived from the production history of DDT (6). A 2-cm sampling interval was applied to the 36 cm deep core 3C1. For the Santa Barbara Basin, 8 horizons in the upper 4.0 cm and 11more horizons to 60 cm were analyzed. Neutron activation analysis of the distilled water washed and dried sediments was carried out a t the MIT nuclear reactor (25). Seven to eight REE concentrations were determined (La, Ce, Nd, Sm, Eu, Tb, Yb, Lu) along with several major elements (AI, Fe, Mn). Precision for La, Ce, Nd, Sm, Eu, Yb, AI, Fe, and Mn is i5-1070. T b and Lu Environ. Sci. Technoi., VoI. 25, NO. 2. 1991
311
La (x), Ce (I), ppm 0
20
40
60 80 100 120 140 160 180 200 220 240 260
2 4 6
8 10 12
14
24 26
28 30 32 34 10
20
70
80
90
100
110
I
50 I
60
I
30 I
40
I
I
I
I
I
I
I
1
2
3
4
5
6
7
8
9
10
11
I
1
% Organic Carbon (o), Hydrocarbons (o), mgg -1 Figure 2. Depth profiles of the San Pedro Shelf core 3C1: La, Ce, organic carbon, DDT, and total hydrocarbons. The three organic constituents come from the same samples ( 1 4 ) as do the REE.
are less precise at &15-30%. Organic carbon, DDT, and hydrocarbon data for core 3C1 are reported in Eganhouse and Kaplan (14). Four sediments from the San Pedro Shelf core were leached with 1 N HC1 to assess the amount of REE released with a mild acid treatment. To 1g of wet sediment in a 50-mL centrifruge tube was added 20 g of 1 N HC1. After being shaken at room temperature for 11h, the tube was centrifuged and the resulting solution was filtered through a Millipore Sterivex-GV syringe filter (0.22-pm pore size). The sediment was quickly rinsed with 20 mL of Milli-Q water. The resulting solution was filtered through the same syringe filter and added to the 1 N HC1 filtrate. The REE composition of this solution was measured by ICP-ES after a separation step (26). Two samples of final effluent particulates from the JWPCP plant were analyzed by INAA. These samples have been frozen since their collection in 1979 by one of us (R.P.E.). Sample PCP was collected on 15 February 1979; after the supernatant liquid was centrifuged and decanted, the solid material was freeze-dried. Sample NCP, collected on 15 May 1979, had its freeze-dried sediment extracted with dichloromethane to isolate the lipids before being frozen. Archived wastewater particles are rare; hence, we only report on these two samples. It should be emphasized that such samples are extremely heterogeneous, containing 60-70'70 organic matter and large amounts of fibrous material. Our samples were not homogenized,but INAA measures the total REE concentration. 4. Results and Discussion
A. San Pedro Basin Shelf. The REE and major element composition of the San Pedro Shelf sediments 312
Environ. Sci. Technoi., Vol. 25, No. 2, 1991
(core 3C1) are presented in Table 11. Figure 2 compares the depth profiles of La and Ce with those of organic carbon, DDT, and hydrocarbons. Figure 3 contrasts profiles of the LREE, MREE, and the HREE, and Figure 4 shows shale-normalized patterns of five horizons. In contrast to the uniform concentrations of MREE and HREE throughout the core, the LREE exhibit large concentration gradients between 20 and 10 cm and high concentrations in the upper 10 cm. The three major elements exhibit no consistent gradients, with Al, Fe, and Mn concentrations being relatively constant in the upper 20 cm (Table 11). Below this depth the variations are probably due to changes in mineralogical composition (Le., proportions of sand, silt, and clay). Iron and especially Mn can be diagenetically active elements due to their oxidation-reduction chemistries (27,28). Many sediments, including those from Santa Monica Basin (19),have elevated Fe and Mn concentrations in near-surface sediment due to redox-driven remobilization and deposition. Core 3C1 does not exhibit these processes in the solid phase. Concentration profiles and shale-normalized patterns show that the deposition of a highly LREE enriched sediment starts above the 18-20-cm horizon. These sediments have unique REE signatures. Deeper sediments have REE abundances similar to crustal material (Figure 4). Peak (2-4 cm) to background (34-36 cm) concentration ratios for La, Ce, Nd, and Sm are 9.6, 3.8, 4.2, and 3.0, respectively. The 2-4-cm horizon has shale-normalized ratios between 6.0 and 1.7 for these four LREE. The concentration decrease in the surface sample (0-2 cm) for the LREE is probably due to a change in the input of LREE-enriched material to core 3C1. Data from a more recent core (3C1 was collected in October 1981) are needed
Table 11. San Pedro Shelf Data 0-2
2-4
4-6
6-8
depth, cm 8-10
170 130 75 8.40 1.10 0.71 2.50 0.39
250 200 110 13 1.30 0.75 2.40 0.46
190 170 86 9.60 1.30 0.72 1.80 0.50
160 182 81 8.40 1.60 0.71 2.30 0.35
140 175 60 7.70 1.30 0.67 2.50 0.32
6.56 4.12 460
6.92 4.39 460
6.60 4.01 450
6.28 4.23 400
18-20
20-22
22-24
24-26
10-12
12-14
14-16
16-18
6.45 3.87 420
120 140 52 7.30 1.30 0.71 2.30 0.38 6.54 4.18 400
52 79 33 6.00 1.10 0.73 2.70 0.56 6.77 4.70 430
86 110 39 6.60 1.20 0.77 2.30 0.48 7.21 4.31 420
47 78 36 5.50 1.20 0.76 2.40 0.39 6.72 4.80 450
deoth. cm 26-28
28-30
30-32
32-34
34-36
1.90 0.29
35 63 29 4.80 1.20 0.66 2.20 0.30
26 52 25 4.30 1.20 0.57 2.30 0.39
29 59 24 4.80 1.20 0.62 2.50 0.36
28 53 25 4.10 1.10 0.51 2.30 0.32
7.41 5.41 455
7.27 5.43 500
5.56 3.91 430
6.40 4.44 470
5.68 3.80 440
33 62 30 5.00 1.30 0.53 2.60 0.39
33 62 29 5.00 1.20 0.43 3.00 0.42
26 53 28 4.30 1.30 0.50 2.90 0.33
27 59 31 4.60 1.30
6.37 3.87 450
5.78 3.57 425
5.83 3.34 450
26 53 26 4.40 1.10 0.49 2.30 0.36 5.98 3.77 410
Table 111. Santa Barbara Basin Data 0-0.25
0.25-0.5
0.5-0.6
1.0-1.5
26.2 57.8 4.19 1.02 0.45 1.73 0.37 5.01 3.27 270
20.0 42.9 3.20 0.82 0.39 1.59 0.14
23.0 56.8 3.79 0.97 0.44 1.87 0.32
5.05 2.47 260
5.23 2.95 280
20.6 48.4 3.27 0.85 0.47 1.50 0.33 5.41 2.57 270
depth, cm 1.5-2.0 2.0-2.5
2.5-3.0
3.0-4.0
4.0-5.0
5.0-6.0
27.3 68.9 4.54 1.13 0.64 2.16 0.38
29.0 63.9 4.91 1.09 0.45 1.69 0.31
26.8 61.3 4.60 1.07 0.80 1.97 0.22
25.0 59.0 4.20 1.08 0.61 1.97
24.3 54.3 3.83 0.96 0.52 1.74 0.41
36.8
5.46 3.53 275
5.40 3.15 275
6.01 3.13 332
5.28 3.62 282
4.90 1.68 298
5.58 3.47 317
5.71 1.24 2.46 1.99 0.28
denth. cm I
Mn, ppm
,
6.0-7.0
8.0-9.0
10-12
14-16
30.7 4.89 1.07 1.72 1.70 0.39 5.14 2.93 306
27.4 4.55 1.03 0.34 1.70 0.19
27.9 4.71 1.01 0.33 1.48 0.20
28.4 4.61 1.03 1.65 1.66 0.25
28.1 5.04 1.00 2.25 1.42 1.06
4.95 3.08 280
5.33 2.94 286
5.38 3.00 297
5.00 2.63 295
to confirm a recent change in the discharge and relative abundance of REE. This explanation is consistent with the sediment profiles and the discharge data of trace metals and organic contaminants in Stull et al. (18). For example, the effluent emissions of Zn from the JWPCP outfalls decreased from 1400 in 1971 to 251 in 1981 to 95 tons/year in 1985. Similarly mass emission rates decreased from 167 000 to 84 000 to 43 000 ton/year. The major feature of our data-high concentrations and enrichment of the LREE in sediments overlying normal concentrations of all the REE-has not previously been observed in cores from other shelf/slope sites where REE concentrations are uniform with depth (26). Pore water studies of coastal reducing sediments indicate that dia-
20-21
24-25 27.6 4.77 0.96 2.28 1.66 0.36 5.21 2.79 287
31-32
41-42
51-52
30.1 4.77 1.06 1.65 1.59 0.94
29.2 4.91 1.06 1.38 1.64 0.16
33.5 5.34 1.21 1.43 1.86 0.77
5.38 2.85 290
5.35 2.59 310
5.64 3.19 316
genetic reactions, though active, do not significantly affect the solid-phase distribution of REE (29, 30). Hence, the major features of core 3C1 cannot be produced by diagenetic reactions. The vertical distributions of the LREE are different from those of organic carbon, DDT, and hydrocarbons (Figure 2), all of which reflect inputs of waste-derived particulate matter. As developed in Eganhouse and Kaplan (14),the gradients in these organic constituents reflect temporal variations in both the input to and output from the wastewater treatment systems. These include (1) production of DDT in 1950 followed by the 1971 banning of its input to the wastewater system and (2) improvements in the sewage treatment processes (14, 18). Before 1946 Environ. Sci. Technol., Vol. 25, No. 2, 1991 313
0
2
San Pedro Shelf
4 6 8
10
2-4 cm 8-10 cm
12
14 16
2
s
18
22
24
1 .o
26
0 0 .
1
La Ce
I 20
40
60
80
100 120
140 160
La (x), Ce (m), Nd
(A),
180
200
220
240 260
ppm
1
Pr
1 Nd
1
1
1
Sm
Eu
1
1
Gd Tb
1 Dy
1
Ho
1 Er
I Tm
I Yb
Lu
RE€ Figure 4. Shale-normalized ratios of selected horizons from Sari Pedro Shelf core 3C 1.
Table IV. S a n Pedro Shelf Core: Ratios of REE Concentratioins of 1 N HCl Leach to Total“
0-2 La Ce Nd Sm
Eu Yb Lu
0.73 (133) 0.65 (85) 0.92 (69) 0.73 (6.1) 0.56 (0.62) 0.35 (0.88) 0.38 (0.15)
depth, cm 6-8 32-34 0.64 0.56 0.67 0.66 0.46 0.35 0.38
0.32 (90) 0.29 (15) 0.35 (8.8) 0.48 (2.0) 0.32 (0.35) 0.35 (0.81) 0.43 (0.14)
34-36 0.27 0.27 0.31 0.41 0.30 0.34 0.37
Data in parentheses represent ppm in sediments derived from 1 N HC1 leach.
34
r
+
B
lo
+
Eu (oh Yb (+A Lu (*), ppm Figure 3. Depth profiles of the REE for the San Pedro Shelf core (3C1). (a) Light rare earth elements-La, Ce, Nd, and Sm. (b) Middle and heavy rare earth elements-Eu, Yb, and Lu.
and 1970 the JWPCP solid waste emissions increased almost linearly from 30000 to 170000 metric tons/year. From this peak value, the emissions decreased to 84 000 in 1981 (14). The organic constituents exhibit subsurface maxima between 6 and 10 cm; the DDT maximum is broader in extending from 6 to 16 cm. The La, Ce, Nd, and Sm peaks are much shallower a t 2-4 cm. From the different gradients it is apparent the input function of the REE differs from the organic constituents. The large increases in the latter from 23 to 17 cm cut across the uniform and background concentrations of the LREE. With respect to the LREE, the increase in their concentrations above background levels occurs in the early 1960s
-
314
Environ. Sci. Technol., Vol. 25, No. 2, 1991
(Figure 2). This is consistent with the historical use of LREE-enriched cracking catalysts by the oil-refining industry (see Introduction). By way of comparison, the La/Eu ratio of shale is 26 while that of a “typical” bastnasite is 298 (Table I). For core 3C1 the La/Eu ratios are 154, 192, and 24 for the 0-2-, 2-4-, and 34-36-cm horizons, respectively. The results for the 1 N HC1 leaches are given in Table IV as ratios to total REE concentrations and as ppm. For the deep horizons (32-34 and 34-36 cm), 1 N HC1 removes between 30 and 40% of all the REE. The 30-40% range is consistent with mild acid leaches of coastal sediments that are not contaminated with the REE (29, 31). In contrast, for the upper sediments (0-2 and 6-8 cm), La, Ce, Nd, and Sm are removed to the extent of 54-90% while the HREE, Yb, and Lu remain a t 35-38% of the total concentration. The LREE-enriched signature is evident in the mildly leached fraction of the upper sediments. While a 1 N HC1 leach will not reveal the specific phases carrying the LREE, our data are consistent with the presence of LREE-enriched zeolites and the products of their oil cracking. For example, the LREE-enriched composition of water percolating through fly ash deposits of oil-fired power plants indicates that these REE are easily leachable (12). Table V includes the REE concentrations and shalenormalized ratios of the final effluent particulates from the JWCPC plant in 1979. Both samples show high concentrations of the LREE and patterns strongly enriched
Table V. REE Concentrations a n d Shale-Normalized Ratios of Final Effluent Particles from the J W P S P Plant in 1979
La Ce Sm Eu Tb Yb Lu
pep, PPm
PCP norm
1050.000 760.000 29.000 0.810 1.900 2.700 0.099
25.60 9.10 3.87 0.50 1.54 0.70 0.16
NCP, P P ~NCP norm 645.000 770.000 38.000 0.680 2.200 2.000 0.075
Santa Barbara 0
15.70 9.30 5.07 0.42 1.79 0.57 0.12
in LREE relative to the crust. La is the most enriched with ratios of 26 and 16. Ce and Sm are enriched by factors of 4-9. For the MREE and HREE the ratios range between 1.5 and less than l. The 0-2-cm sediments of core 3C1, by way of comparison, have enrichments of La, Ce, and Sm of 6.1, 2.4, and 1.7, respectively. These data demonstrate that the effluent released in 1979 through the JWCPC plant outfall has the noncrustal signature observed in the upper sediments of the San Pedro Shelf. B. Santa Barbara Basin. As seen in Figure 5 and Table 111, all the REE in the Santa Barbara Basin show similar vertical distributions with no indication of any enrichment of LREE in the upper horizons. The upper 20 cm corresponds to about the past 40 years. The REE and Al, Fe, and Mn concentrations show sawtoothlike variations in the upper 8 cm followed by smoother profiles to 60 cm. The shale-normalized patterns for the Santa Barbara Basin core are nearly flat (data not shown). Hence, the REE signatures of the Santa Barbara Basin sediments and the deeper “background” sediments of the San Pedro Shelf are similar to crustal material. Conclusions
Our data indicate that refined petroleum products from Los Angeles are responsible for the LREE-enriched signature in the upper 20 cm of the San Pedro shelf core. As discussed by Olmez and Gordon (111, LREE-containing catalysts (1-3% as oxides) are used in the (Los Angeles) oil refineries for the cracking of petroleum. According to Preinfalk and Morteani (7), this industrial use represents about 1/3 of the REE consumption in the United States; metallurigcal and ceramic/glass industries each use Olmez and Gordon (11)estimated that 2000 tons of catalyst material is lost per day by U.S. oil refineries to the atmosphere with cracking products such as fuel oil. The dominant minerals in REE deposits, bastnasite and monazite, are characterized by large enrichments in La, Ce, Nd, and Sm (8-10). Bastnasite is a fluorcarbonate mineral and monazite is a phosphate mineral. The date in Table I for these minerals show that La, Ce, Nd, and Sm are 890-6900 times enriched over crustal material. In the metallurgical and glass industry, individual REE are used for different processes and products (8). Hence, their anthropogenic signature will be very different from cracking catalysts that use ores enriched in all the LREE. The anthropogenic input of LREE to the San Pedro Shelf site corresponds to the early and mid-1960s or the 12-18-cm depth zone where LREE concentrations begin to increase over background values. The location of core 3C1 and the high concentration of LREE in the upper sediments indicate that the anthropogenic component of the LREE is being deposited from the JPWCP wastewater outfall. Our data for two particulate effluent samples confirm this. This outfall is the major injection point for anthropogenic components of trace metals (4, 18-20), organic matter (241, and trace organic compounds (14) found
I
I
08
10
12
1.4
Eu
16
18
20
1
22
(Oh Yb (+h ppm
Flgure 5. Sediment profiles of REE from the Santa Barbara Basin core-La, Sm, Eu, and Yb.
a t this same San Pedro Shelf site. By analogy with other trace (heavy) metals (20),the largest proportion of REE discharged from the outfalls probably occurs in the particulate form, although data specific to the phase associations of effluent REE do not exist. There are no previously published REE data for the southern California Borderland system (Le., seawater, particles, atmospheric particles, and wastewater discharges). While data from one core and two effluent particulate samples clearly indicate an anthropogenic source to the San Pedro Shelf, additional data would be required to understand the flux, redistribution, and signature of REE in this region, which is characterized by intense and varied anthropogenic sources. The unique signature and source of REE in the San Pedro Shelf core and in the JWPCP discharge offers yet another tool for tracing wastewater discharges. Critical to a more detailed study would be data on the past and current REE composition of wastewater discharges, atmospheric particles, and recently collected cores in the San Pedro and Santa Monica Basins. Acknowledgments
Dr. C. Reimers, Scripps Institution of Oceanography kindly provided us with subsamples of her Santa Barbara Basin core. Sediments from San Pedro Shelf were provided by J. Stull and T. Hessen of the Los Angeles County Sanitation District. At WHO1 Ed Brook and Julianne Palmieri carried out analyses of the sediment leachates. Registry No. La, 7439-91-0; Ce, 7440-45-1; Nd, 7440-00-8; Sm, 7440-19-9; Eu, 7440-53-1; Tb, 7440-27-9; Yb, 7440-64-4; Lu, 7439-94-3; Al, 7429-90-5; Fe, 7439-89-6; Mn, 7439-96-5.
Literature Cited (1) Henderson, P. In Rare Earth Element Geochemistry; Henderson, P. Ed.; Elsevier: Amsterdam, 1984; p p 1-32. Environ. Sci. Technol., Vol. 25, No. 2. 1991 315
Environ. Sci. Technol. 1991, 25, 316-322
Taylor, S. R.; McLennan, S. M. The Continental Crust: Its Composition and Euolution; Blackwell: Oxford, 1985. Elderfield, H. Philos. Trans. R. SOC.London 1988, A325, 105-126. Katz, A.; Kaplan, I. R. Mar. Chem. 1981, 10, 261-299. Plank, C. J. CHEMTECH 1984, 243. Pines, H. The Chemistry of Catalytic Hydrocarbon Conuersions; Academic Press: New York, 1981. Preinfalk, C.; Morteani, G. In Lanthanides, Tantalum, and Niobium; Moller, P., Cerny, P., Saupe, F., Eds.; Springer-Verlag: Berlin, 1989; pp 359-370. Cesbron, F. P. In Lanthanides, Tantalum, and Niobium; Moller, P., Cerny, P., Saupe, F., Eds.; Springer-Verlag: Berlin, 1989; pp 3-26. Neary, C. R.; Highley, D. E. In Lanthanides, Tantalum, and Niobium; Moller, P., Cerny, P., Saupe, F., Eds.; Springer-Verlag: Berlin, 1989; pp 423-466. Moller, P. In Lanthanides, Tantalum, and Niobium; Moller, P., Cerny, P., Saupe, F., Eds.; Springer-Verlag: Berlin, 1989; pp 171-188. Olmez, I.; Gordon, G. E. Science 1985, 229, 966-968. Olmez, I. In Receptor Models in Air Resources Management; Watson, J. G., Ed.; APCA: Pittsburg, PA, 1989;pp 3-11. Olmez, I.; Sheffield, A. E.; Gordon, G. E.; Houck, J. E.; Pritchett, L. C.; Cooper, J. A.; Dzubay, T. G.; Bennett, R. L. J . Air. Pollut. Control Assoc. 1988, 22, 1392-1402. Eganhouse, R. P.; Kaplan, I. R. Mar. Chem. 1988, 24, 163-191. Chow, T. J.; Bruland, K. W.; Bertine, K.; Soutar, A.; Koide, M.; Goldberg, E. D. Science 1973, 181, 551-552. Bruland, K. W.; Bertine, K.; Koide, M.; Goldberg, E. D. Environ. Sei. Technol. 1974, 8, 425-432. Bertine, K. S.; Goldberg,E. D. Enuiron. Sci. Technol. 1977, 11, 297-299. Stull, J. K.; Baird, R. B.; Heesen, T. C. J.-Water Pollut. Control Fed. 1986, 58, 985-991.
(19) Finney, B. P.; Huh, C.-A. Enuiron. Sei. Technol. 1989,23, 294-303. (20) Galloway, J. N. Geochim. Cosmochim. Acta 1982, 46, 2307-2321. (21) Koide, M.; Soutar, A,; Goldberg, E. Earth Planet. Sci. Lett. 1972, 14, 442-446. (22) Bruland, K. W.; Franks, R. P.; Landing, W. M. Earth Planet. Sei. Lett. 1981, 53, 400-408. (23) Stull, J. K.; Haydock, C. I.; Montagne, D. E. Estuarine, Coastal Shelf Sei. 1986, 22, 1-17.
(24) Schwdbach, J. R.; Gorsline, D. S.J. Sediment. Petrol. 1985, 55, 829-842. (25) Germani, M. S.; Gokmen, I.; Sigleo, A. C.; Koiwalczyk, G. S.; Olmez, I.; Small, A. M.; Anderson, D. L.; Failey, M. P.; Gulovali, M. C.; Choquette, C. E.; Lepel, E. A.; Gordon, G. E.; Zoller, W. H. Anal. Chem. 1980, 52, 240. (26) Sholkovitz, E. R. Am. J . Sci. 1988, 288, 236-281. (27) Trefry, J. H.; Presley, B. J. Geochim. Cosmochim. Acta 1982, 46, 1715-1726. (28) Burdige, D. J.; Gieskes, J. M. Am. J . Sei. 1983,283, 29-47. (29) Elderfield, H.; Sholkovitz, E. R. Earth Planet. Sei. Lett. 1984, 82, 280-290. (30) Sholkovitz, E. R.; Piepgras,D. J.; Jacobsen, S. B. Geochim. Cosmochim. Acta 1989, 53, 2847-2856. (31) Sholkovitz, E. R. Chem. Geol. 1989, 77, 47-51. Received for reuiew December 18, 1989. Revised manuscript received June 19,1990, Accepted October I O , 1990. E.R.S. was supported by NSF Grant OCE-85-15695. This is contribution No. 7264 from the Woods Hole Oceanographic Institution. Special thanks to the MITR-11 reactor staff for their assistance during irradiations. This work was supported in part by the U S . Department of Energy under Grant DE-FG-2-80ER10770. However, any opinions, findings, conclusions, or recommendations expressed herein are those of the authors and d o not necessarily reflect the view o f DOE.
Heterogeneous Photolysis of Polychlorinated Dibenzo-p -dioxins on Fly Ash in Water-Acetonitrile Solution in Relation to the Reaction with Ozone Hajime Muto”
Environmental Research Center, Akita University, Hondo 1-1-1, Akita 010, Japan Masayuki Shinada and Yukio Takkawa
Department of Public Health, Akita University School of Medicine, Hondo 1-1-1, Akita 010, Japan The destruction of polychlorinated dibenzo-p-dioxins (PCDDs) on fly ash particles from municipal solid waste in water or water-acetonitrile (2:3, v/v) solution was conducted by heterogeneous photolysis (254 nm) with or without the presence of ozone. The photolytic reactions of 2,3,7,8-chlorine-substitutedDDs of tetra- through octa-CDDs were faster than those of other isomers in several photosystems. The presence of ozone facilitated PCDD photolyses as well as the “PCDD risk rate” on fly ash. Fly ash particles themselves, which include PCDDs and metal oxides, play a role in the photodriven process for PCDD destruction in the heterogeneous photosystems. The toxic equivalent calculation is necessary to assess the effect of PCDDs on human health and to develop an advanced technique for PCDD reduction. Introduction Recently, the formation, decomposition, and/or environmental fate of polychlorinated dibenzo-p-dioxins (PCDDs) and polychlorinated dibenzofurans (PCDFs) by incineration of municipal solid wastes (MSWs) has been 316
Environ. Sci. Technol., Vol. 25, No. 2, 1991
discussed from the point of view of human health effects (1-6), and risk assessment and management studies on these compounds have been conducted in various countries (7-9). Several studies on the physical properties, e.g., solubilities or molar absorptivities, of PCDDs and on their liquid, soil, or catalytic photolyses have been reported (10-23). However, the study of heterogeneous photolysis with fly ash particles has not been reported. In order to clarify the environmental fate of PCDDs and to develop a treatment technique for PCDDs on fly ash, which were accumulated by incineration, we report here the heterogeneous photolyses, with or without the presence of ozone, of PCDDs on fly ash in water or water-acetonitrile solution, using GC/MS analyses. Materials a n d Methods Reagents. n-Hexane, acetonitrile, benzene, dichloromethane, acetone, hydrochloric and sulfuric acids, and alumina were reagent grade quality and obtained from Wako Pure Chemical Industries (Osaka, Japan). PCDD standards of 2,3,7,8-tetra-CDD, 1,2,3,4-tetra-CDD,pentathrough hepta-CDDs, 1,2,3,4,6,7,8,9-octa-CDD,and PCDD
0013-936X/91/0925-0316$02.50/0
8 1991 American Chemical Society
