Environ. Sci. Technol. 2010, 44, 3876–3882
Evaluating the Behavior of Gadolinium and Other Rare Earth Elements through Large Metropolitan Sewage Treatment Plants P H I L I P L . V E R P L A N C K , * ,† EDWARD T. FURLONG,‡ JAMES L. GRAY,‡ PATRICK J. PHILLIPS,§ RUTH E. WOLF,| AND KATHLEEN ESPOSITO⊥ U.S. Geological Survey, MS 973, Denver Federal Center, Denver, Colorado 80225, U.S. Geological Survey, MS 407, Denver Federal Center, Denver, Colorado 80225, U.S. Geological Survey, 425 Jordan Road, Troy, New York 12180, U.S. Geological Survey, MS 964, Denver Federal Center, Denver, Colorado 80225, and AECOM, 605 3rd Avenue, New York, New York
Received December 22, 2009. Revised manuscript received April 2, 2010. Accepted April 5, 2010.
A primary pathway for emerging contaminants (pharmaceuticals, personal care products, steroids, and hormones) to enter aquatic ecosystems is effluent from sewage treatment plants (STP), and identifying technologies to minimize the amount of these contaminants released is important. Quantifying the flux of these contaminants through STPs is difficult. This study evaluates the behavior of gadolinium, a rare earth element (REE) utilized as a contrasting agent in magnetic resonance imaging (MRI), through four full-scale metropolitan STPs that utilize several biosolids thickening, conditioning, stabilization, and dewatering processing technologies. The organically complexed Gd from MRIs has been shown to be stable in aquatic systems and has the potential to be utilized as a conservative tracer in STP operations to compare to an emerging contaminant of interest. Influent and effluent waters display large enrichments in Gd compared to other REEs. In contrast, most sludge samples from the STPs do not display Gd enrichments, including primary sludges and end-product sludges. The excess Gd appears to remain in the liquid phase throughout the STP operations, but detailed quantification of the input Gd load and residence times of various STP operations is needed to utilize Gd as a conservative tracer.
Introduction A primary pathway for emerging contaminants (pharmaceuticals, personal care products, steroids, and hormones) to enter aquatic ecosystems is effluent from municipal sewage treatment plants (STP (1)). These complex organic compounds have a wide range of physicochemical, structural, * Corresponding author phone: (303)236-1902; fax: (303)236-3200; e-mail:
[email protected]. † U.S. Geological Survey, MS 973. ‡ U.S. Geological Survey, MS 407. § U.S. Geological Survey, Troy, NY. | U.S. Geological Survey, MS 964. ⊥ AECOM. 3876
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 44, NO. 10, 2010
and electrostatic characteristics, such that their behavior during STP operations is quite variable. Furthermore, the behavior of many contaminants depends on which wastewater treatment operations are utilized at a given facility and factors such as chemical reactions, biodegradation, ion exchanges, electrostatic attractions, disinfection, and variations in pH, dissolved oxygen, temperature, and pressure. Identifying STP operations that decrease the mass of emerging contaminants released to the environment is difficult because of the complexity of the wastewater stream and the variability of the input mass. Figure 1 displays a schematic diagram showing the range of typical wastewater and biosolids treatment processes. Approaches that have been used to evaluate the extent of removal of emerging contaminants include the mass balance approach to quantify the flux of constituents of interest in the liquid and solid phases of constituents of interest (2, 3) and an estimation of the fraction removed in the solid phase by trying to determine partition coefficients of the transfer from the liquid to the solid phases (2, 4, 5). Quantitative assessment also is difficult because of the uncertainty involved. Although analytical uncertainty has been greatly improved in recent years, uncertainty from other sources including inflow variability, estimation of percent change in each step, and variability in residence time of each operation along the treatment train still remains. In this study we evaluated the behavior of gadolinium and other rare earth elements (REE) during the operation of large metropolitan STPs in the United States. Identifying constituents that remain in the aqueous phase throughout STP operations indicates their potential for use as conservative tracers. Quantifying the extent of removal of emerging contaminants can then be undertaken by comparing emerging contaminant concentrations to a conservative tracer. Gadolinium was chosen for the focus of this study because recent work has shown that anthropogenic Gd is quite enriched in STP effluent and is quite stable in rivers receiving effluent from STPs (6, 7). During the past 10 years, gadolinium has been documented to be enriched, relative to its neighboring REEs europium and terbium, in STP effluent in large urban areas (6-12). The enrichment in STP effluent has been traced to the use of gadolinium organic complexes as contrasting agents for magnetic resonance imaging (MRI). These Gd organic complexes are either linear compounds (Gd-DTPA, Gd-DTPA-BMA, and Gd-BOPTA) or macrocyclic compounds (Gd-DOTA and Gd-BT-DO3A) (13). Contrasting agents are used to enhance images in patients undergoing MRIs and are taken either orally or intravenously (14). Gadolinium is an ideal element to use as a contrasting agent because it has a large magnetic dipole moment due to multiple unpaired electrons and, as an organic complex, is quite inert; it is excreted through the kidneys with a half-life of 2 h. Organic Gd compounds are extremely stable, and traditional STP practices may not break them down (11). Recent studies have documented the persistence of the Gd enrichment downstream from a STP for at least 14 km and also through estuaries into coastal areas (6, 7). Organically complexed, medical waste derived Gd is not the only source of Gd to municipal wastewater treatment plants. Rare earth elements, including Gd, are found in nature as individual REE mineral phases or as minor or trace elements in other mineral phases. The weathering of rocks and minerals leads to the mobilization of REEs and the presence of REEs in water as dissolved species or associated with colloids. In general, in natural systems gadolinium is not enriched compared to its neighboring REEs because of 10.1021/es903888t
2010 American Chemical Society
Published on Web 04/16/2010
FIGURE 1. Schematic diagram showing the range of typical wastewater and biosolids treatment processes. their similar chemical properties, and since it is not present as a MRI-derived Gd organic complex should behave similarly to the other REEs. Although conservative behavior of anthropogenic Gd downstream from STPs is well documented, only limited research on the behavior of Gd within STPs has been undertaken. Verplanck et al. (6) report shale-normalized REE data for a biosolid (the composited, solid phase material from STPs) from a large, urban STP in western United States and document that no Gd enrichment was found even though a large positive Gd anomaly was found in the STP effluent. Vivian (1986) (15) also reported no Gd anomaly in sewage sludge from the Liverpool U.K., but this study was conducted prior to the widespread use of Gd organic complexes as contrasting agents. Ku ¨ nnemeyer et al. (13) developed a methodology for simultaneous speciation analysis of Gd organic complexes and tested the methodology with a suite of water samples collected over a 1-h period from the STP in Mu ¨ nster Germany. They concluded that the lower concentrations of some of the Gd organic compounds through STP operations were evidence that the Gd compounds were unstable. Lawrence and Ort (16) wrote a comment on the manuscript arguing that the authors overstated their conclusion because the sample collection was insufficient to capture the potential dynamic change in loads through the STP. The objective of this study was to evaluate the behavior of Gd and other REEs through large, metropolitan STP operations. Four full-scale metropolitan STPs that utilize several biosolids thickening, conditioning, stabilization, and dewatering processing technologies were sampled. This survey was designed to evaluate if any solid-phase extraction techniques had the potential to remove a substantial fraction of Gd compared to other REEs. Solid phase samples of sludges generated at each stage of treatment were collected. In addition, at two STPs influent and effluent aqueous samples were collected. The liquid samples were collected over a 24-h period, and primary and secondary effluent sampling was delayed to try to match the travel time through the treatment plant. Solid samples were collected during the same time period and, if necessary, at other times. For one STP solid and liquid chemical results are integrated into a mass-flow balance to determine whether the mass of Gd removed from the liquid phase would be substantial enough to produce an
TABLE 1. Study Site Informationa
capacity population served service area began operating a
Plant 1
Plant 2
Plant 3
Plant 4
1740 MLD 9,000,000 1330 km2 1894
680 MLD 994,000 360 km 2 1907
34 MLD 60,000 140 km2 NA
1400 MLD 2,000,000 NA 1938
MLD (millions of liters per day), NA (not available).
enrichment in Gd in the rare earth element pattern of the solid phase.
Methods Solid phase samples from four STPs that serve large metropolitan areas in the United States were collected between December 2005 and February 2007 (Table 1). Solid samples were collected as a series of grab samples over a period of a few hours. Because of on going STP operations, a 24 h composite sampling was not possible. This suite of samples covered a wide-range in treatment-plant derived solids including primary unthickened and thickened sludge, dewatered sludge, anaerobically digested sludge, acid phase digested sludge, methane phase digested sludge, thickened secondary sludge, tertiary pelletized sludge, and tertiary composted sludge. In general, municipal sewage is treated in three stages: a coarse screening to remove large solids, a chemical treatment to promote flocculation, and a biological step to promote degradation of organic material. Various sludges are produced at each step and then blended, thickened, and dewatered to produce end-product sludges (2-4). Aliquots of solid phase samples for inorganic chemical analyses were taken from the composite samples and airdried. Dried samples were powdered using an agate mortar, weighed, and ashed at 500 °C for 13 h. Ashed samples were reweighed and digested with a mixture of hydrochloric, nitric, perchloric, and hydrofluoric acids (17). At two STPs, composite water samples were collected of STP influent, primary effluent, and secondary effluent. Liquid samples were collected as 24-h, flow-weighted composites: hourly 0.75 L samples were collected and volumes withdrawn from each sample based on the flow for that hour. The start of the 24-h sampling was lagged for the primary and VOL. 44, NO. 10, 2010 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
3877
TABLE 2. Rare Earth Element Concentrations in Waters in pmol/L, Gd/Gd* Is Unitless According to Eq 1a sample
P1-PIN
P1- PEF
P1- SEF
P2-PIN
P2-PIN
P2-PEF
P2-SEF
date La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu ΣREE Gd/Gd*
6/2006 87.1 239 13.5 47.1 21.3 6.58 1300 2.52 10.5 4.24 9.57 2.37 6.36 2.86 1760 79.0
6/2006 84.9 266 22.7 67.2 14.6 9.87 1670 3.78 11.7 6.06 16.1 3.55 10.0 5.14 2190 87.2
6/2006 56.2 196 10.6 39.5 11.3 2.63 1520 1.89 12.3 3.64 10.2 2.96 15.0 1.14 1880 137
7/2006 391 596 126 420 124 37.5 2050 30.2 134 35.8 79.5 27.8 86.1 32.0 4170 13.2
1/2007 324 457 93.0 424 102 23.7 2090 19.5 99.1 26.1 65.2 8.88 75.7 15.4 3820 19.2
1/2007 153 224 44.0 191 37.9 11.8 1830 11.3 58.5 13.9 37.1 8.29 46.8 10.9 2680 33.5
1/2007 38.2 74.2 14.2 47.1 14.0 5.26 1010 3.78 18.5 5.46 19.7 5.33 27.2 6.86 1290 53.7
a
P1 (Plant 1), P2 (Plant 2), PIN (primary influent), PEF (primary effluent), SEF (secondary effluent).
secondary effluent samples to try to match the hydraulic residence times of the various STP operations, trying to follow a 24-h slug of water through the STP. Primary effluent is the liquid from the primary treatment operation, usually settling, that then goes to the secondary treatment operation. Secondary effluent is the liquid that emerges from the secondary treatment operation (Figure 1). These liquid samples were filtered on site using 0.45 µm pore-size tortuous path, filter membrane and acidified with ultrapure, concentrated HNO3 to a pH of
