Environ. Sci. Technol. 1992, 26, 538-545
Cameron, D. R.; Klute, A. Water Resour. Res. 1977,13,183. Curtis, G. P.; Roberts, P. V.; Reinhard, M. Water Resour. Res. 1986, 22, 2059. Rao, P. S. C.; Davidson, R. E.; Jessup; Selim, H. M. Soil Sci. SOC.A m . J . 1979, 43, 22. Steinberg, S. M.; Pignatello, J. J.; Sawhney, B. L. Environ. Sci. Technol. 1987, 21, 1201. Weber, W. J., Jr.; Miller, C. T. Environ. Sci. Technol. 1988, 22, 457. Ahlert, W. K.; Uchrin, C. G. J . Hazard. Mater. 1990,23, 317. Garbarini, D. R.; Lion, L. W. Environ. Sci. Technol. 1986, 20, 1263. Stauffer, T. B.; MacIntyre, W. G.; Wickman, D. C. Enuiron. Toxicol. Chem. 1989, 8 , 845.
(36) Schwarzenbach, R. P.; Westall, J. Environ. Sci. Technol. 1981,15, 1360. (37) Stauffer, T. B.; MacIntyre, W. G. Enuiron. Toxicol. Chem. 1986, 5 , 949. (38) Elzerman, A. W.; Coates, J. T. In Sources and Fates of Aquatic Pollutants; Advances in Chemistry 216; Hites, R. A., Eisenreich, S. J., Eds.; American Chemical Society: Washington, DC 1987; Chapter 10. Received for review May 9, 1991. Revised manuscript received October 3,1991. Accepted October 18,1991. This research was supported by the New York State Center for Hazardous Waste Management (Contract 150- W009C-Rl0195)and by Chemical Waste Management, Inc.
Herbicide Transport in Rivers: Importance of Hydrology and Geochemistry in Nonpoint-Source Contamination Paul J. Squillace*~tand E. M. Thurmanz US. Geological Survey, Water Resources Division, Iowa City, Iowa 52244 and Lawrence, Kansas 6 6 0 4 9
phenyl)-N-(methoxymethyl)acetamide]for corn and soybeans, atrazine [2-chloro-4-(ethylamino)-6-isopropylamine-s-triazine] for corn and sorghum, cyanazine [2-[[4chloro-6- (ethylamino)-s-triazin-2-yl]amino] -2-methylpropionitrile] for corn, metolachlor [ 2-chloro-N-(2-ethyl6-methylphenyl)-N-(2-methoxy-l-methylethyl)acetamide] for corn and soybeans, and metribuzin [4-amino-6-(1,1dimethylethyl)-3-(methylthio)-1,2,4-triazin-5(~-one] for soybeans. These compounds enter surface water and groundwater through various pathways, and their contribution to nonpoint-source contamination is poorly understood. Research on small fields indicates that losses of herbicides are approximately 1-4%, depending on the tillage practice and slope of the fields (2-8). Furthermore, most herbicide transport occurs within a critical period of 2-6 weeks after application (8,9). Past studies generally have been limited to small fields or occasionally to slightly larger areas (10-100 km2). For example, Frank and Sirons (10) measured atrazine runoff in Quebec, ON, Canada, from areas of 18-79 km2 and found from 0.3 to 1.9% loss on an annual basis. Smaller losses were from sandy soil, and greater losses were from clayey soil. The mean value of atrazine loss for all watersheds in their study was slightly less than 1%.Frank et al. (11)reported that atrazine and Introduction simazine were the only herbicides detected year round in Nonpoint-source (NPS) contamination is a major streamwaters of the area. Muir et al. (12) estimated that water-quality concern in the midwestern United States 1.7% of the applied atrazine was lost from areas of 22-129 according to a recent feature article in ref 1. The article km2 in Quebec, Canada. Glotfelty et al. (8) found that states, “NPS pollution is a principal source of water-quality 2-3% of the atrazine applied in the Wye River drainage problems ... and that agricultural activities are a major basin ( 10 km2)was lost in runoff to Chesapeake Bay, and cause”. loss occurred mainly during the 2-week period immediately In the midcontinent area, especially the states of Illinois, after application. They reported that the loss of atrazine Indiana, Iowa, Kansas, Minnesota, Missouri, Nebraska, decreased substantially within 4-6 weeks after application. and Ohio, herbicides are used extensively to control weeds Much less is known about the fate of herbicides in large that affect corn (Zea mays[L.]), sorghum (Sorghum bicodrainage areas (>500 km2) and whether the transport of lor[L.]), and soybeans (Glycine max[L.]). The result is herbicides in large areas is similar to results of field-disthat large quantities of herbicides (1-3 kg/ha) are used sipation studies. Furthermore, little is understood conannually on crops in these states. The major herbicides cerning the role of groundwater in contributing herbicides include the following alachlor [2-chloro-N-(2,6-diethyl- to surface-water flow. The objectives of this study were (1)to measure the mass of atrazine and several other major herbicides transported from the Cedar River basin in +US.Geological Survey, Iowa City, IA 52244. Minnesota and Iowa (Figure 1)and, in so doing, to improve * U.S. Geological Survey, Lawrence, KS 66049. Alachlor, atrazine, cyanazine, metolachlor, and metribuzin were measured at six sites during 1984 and 1985 in large subbasins within the Cedar River, IA. A computer model separated the Cedar River discharge hydrograph into groundwater and overland-flow components. The concentration of herbicides in the river when groundwater was the major flow component was less than 1.0 pg/L and averaged 0.2 pg/L. The maximum concentrations of herbicides occurred when overland flow was the major component of river discharge, exceeding 50 pg/L for total herbicides. About 6% of the annual river load of atrazine was transported with the groundwater component, while 94% was transported with overland flow. From 1.5 to 5% of the atrazine applied during the year was transported from the basin. Atrazine concentrations in the river increased according to the discharge divided by the drainage area. This correlation indicates that rivers with large normalized 2-year peak flows have the potential to transport large concentrations of herbicides. A diagrammatic model of nonpoint-source transport of herbicides was developed that suggests that sorbed transport from fields occurs during episodes of overland flow with rapid dissolution of herbicides downstream. 4
N
538
Environ. Sci. Technol., Vol. 26, No. 3, 1992
Not subject to U.S. Copyright. Published 1992 by the American Chemical Society
MINNESOTA
Index map
EXPLANATION (6) 775 km2v
SURFACE-WATER-QUALITY SAMPLING SITE--Upper number is the site identification number; lower number is drainage basin size, in square kilometers I k m 1
Figure 1. Study area and sampling sites for Cedar River, Minnesota and Iowa.
the understanding of the role of hydrology (groundwater and overland flow) and geochemistry (sorption and suspended transport) in nonpoint-source contamination on a large scale and (2) to separate the transport of herbicides in the Cedar River between groundwater flow and overland flow and to calculate the amount of herbicide contributed by each of the two sources.
Experimental Section Water samples were collected monthly a t six waterquality sampling sites (Figure 1)from May 1984 through November 1985, and weekly during the period of maximum river discharge (June 1984). All sites were visited during the same 3-day period each month. The discharge at each site was assumed to be equal to the discharge a t a nearby U.S. Geological Survey continuous-record streamflow-gauging station (Figure 1). Because the drainage area of each sampling site was within 95% of the drainage area of a nearby gauging station, the errors introduced by this assumption were *5 % Water samples were collected in 1-L baked-glass bottles in Teflon-lined caps. Samples were centrifuged a t 4000 rpm a t the Iowa Hygienic Laboratory (Iowa City, IA). Both whole-water (total) and centrifuge samples (dis-
.
solved) were analyzed for herbicides. One-liter samples were extracted three times in a separatory funnel with 100 mL of a mixture of methylene chloride and hexane (45% /55% ,v/v). Extracts were combined and evaporated to 3 mL in a Kuderna-Danish evaporator. Atrazine was analyzed by gas chromatography with a nitrogen-phosphorus detector using a packed column of 3% OV-l,lOO-200-mesh gas chrom Q. Before March 1985, metolachlor, alachlor, cyanazine, and metribuzin were analyzed by dual glass-packed columns, which were 1.8 m long with a 2-mm inside diameter. One column was packed with 3% OV-1 on gas chrom Q, whereas the other was packed with mixture of OV-1 and OV-210 at a ratio of 2 to 3. After March 1985, the compounds were analyzed by dual-column detection on a 30-m capillary column of DB-5 and a 30-m DB-1701 column with an electron-capture detector. For both methods, the identity of the herbicides was confirmed by comparison of retention times on both columns. On two occasions, several samples were split and sent to the U.S. Geological Survey laboratory in Arvada, CO, to confirm the analyses completed by the Iowa Hygienic Laboratory. The method consisted of dual-column confirmation of the herbicides with a nitrogen-phosphorusspecific detector (13). Concentrations of atrazine reported Environ. Sci. Technol., Vol. 26, No. 3, 1992 539
C f D A R RIVER BASIN
1
U a y , June. and July 1984 and 1985
2
0
7
I
I
I
0.10
0 08 0.06
c
t
0.04
HUDSON .CHARLES CITY (1961 to 1986 RECORD)
IONIA
I
NORTHWOOD, CONESVILLE
002' 100
,
1,000
\
I
10,000
I 100.000
DRAINAGE AREA, IN SQUARE KILOMETERS
Figure 3. Peak discharge per unit area for 2-year flood versus drainage area for period of record, 1955-1986.
I
-4 0
-35
-3.0
(DISCHARGE,
(
-2.5
-2.0
IN CUBIC METERS PER SECOND
DRAINAGE AREA, IN SOUARE 2 ppm) because the organic carbon content is 1-5%, and the moisture content is less than 20%; most of the herbicide is sorbed to soil (95% sorbed using Kd calculation, eq 4). If rainfall is intense, then there is a mixture of sediment and water transported from the field. As the concentration of suspended sediment decreases to less than 50 000 mg/L, then atrazine desorbs into the dissolved phase (50% sorbed). This concept explains the conflicting results in the literature, where some studies report that atrazine at field edge is primarily dissolved ( 2 , 3 , 6 , 8 ) whereas , others report that as much as 40% of the atrazine is in the solid phase (4,29). It has also been reported that on field-sized plots as much as 58% of metolachlor can be transported on the suspended sediment (30). Equilibrium calculations demonstrate that as the suspended sediment decreases to 10000 mg/L the atrazine will desorb from the sediment into the dissolved phase (10% sorbed). These calculations are substantiated by data collected by Johnson and Baker (31),who studied a 5505-ha agricultural basin in the Cedar River basin. They found that 12% of the atrazine transported in the stream was carried on the suspended sediment when the suspended-sediment concentrations in the stream were 12 320 mg/L. As the concentrations of suspended sediment decreased in the river, a new equilibrium was rapidly achieved. Desorption of atrazine approached 75% of equilibrium values within 3-6 min (32). This “spring-flush” model (Figure 5 ) is consistent with dissolved transport of herbicides and, at the same time, relates herbicide concentrations with TOC, suspended sediment, and overland flow. Yet, the actual transport measured at our sampling sites in large drainage basins indicates only dissolved transport because sediment concentrations are small (30-700 mg/L). Therefore, desorption presumably has occurred. The correlations of the herbicide concentrations in Table I1 show that significant correlations exist among the concentrations of atrazine and metolachlor (0.94), atrazine and cyanazine (0.94), and atrazine and alachlor (0.80); there is a nonsignificant correlation of atrazine with metribuzin (0.60). These findings may be understood as the correlation of use patterns for the various herbicides. The most extensively used herbicides for corn in a crop-reporting 544
Environ. Sci. Technol., Vol. 26, No. 3, 1992
district in the basin are atrazine, alachlor, and metolachlor, with lesser use of cyanazine. Metribuzin, on the other hand, is used on soybeans. Finally, the correlation of log atrazine concentration with log Q / A (Figure 2) indicates that overland flow is an important mechanism for increased herbicide concentration in rivers and that streams susceptible to large concentrations of atrazine may be identified by analysis of the 2-year flood peaks. Furthermore, a diagrammatic model of sorbed transport from the edge of the field is linked to overland flow and rapid desorption of herbicide in the stream to explain the dynamics of nonpoint-source contamination by herbicides. Registry No. Alachlor, 15972-60-8; atrazine, 1912-24-9; cyanazine, 21725-46-2;metolachlor, 51218-45-2;metribuzin, 2108764-9.
Literature Cited (1) Humenik, F. J.; Smolen, M. D.; Dressing, S. A. Enuiron. Sci. Technol. 1987,21, 737-742. (2) Hall, J. K.; Pawlus, M.; Higgins, E. R. J . Enuiron. Qual. 1972, 1, 172-176. (3) Hall, J. K. J. Enuiron. Qual. 1974, 3, 174-180. (4) Ritter, W. R.; Johnson, H. P.; Lovely, W. G.; Molnau, M. Environ. Sci. Technol. 1974, 8, 38-42. (5) Triplett, G. B., Jr.; Conner, B. J.; Edwards, W. M. J. Enuiron. Qual. 1978, 7, 77-84. (6) Leonard, R. A.; Langdale, G. W.; Fleming, W. G. J. Environ. Qual. 1979, 8, 223-229. (7) Rhode, W. A,; Asmussen, L. E.; Hauser, E. W.; Hester, M. L.; Allison, H. D. Agro-Ecosystems 1981, 7, 225-238. (8) Glotfelty, D. E.; Taylor, A. W.; Isensee, A. R.; Jersey, J.; Glen, S.J. Enuiron. Qual. 1984, 13, 115-121. (9) Wauchope, R. D. J . Enuiron. Qual. 1978, 7, 459-472. (10) Frank, R.; Sirons, G. J. Sci. Total Environ. 1979, 12, 223-229. (11) Frank, R.; Braun, H. E.; Van Hove Holdrinet, M.; Sirons, G. J.; Ripley, B. D. J . Enuiron. Qual. 1982, 11, 497-505. (12) Muir, D. C. G.; Yoo, J. Y.; Baker, B. B. Arch. Environ. Contam. Toxicol. 1978, 7 , 221-235. (13) Wershaw, R. L.; Fishman, M. J.; Grabbe, R. R.; Lowe, L.
E. Techniques of Water-Resources Investigations, Book (14)
(15) (16) (17) (18)
(19) (20) (21) (22)
(23) (24)
5 , Chapter A3; U S . Geological Survey: Reston, VA, 1987; pp 1-80. Skougstad, M.; Fishman, M. J.; Friedman, L. C.; Erdmann, D. E.; Duncan, S. S. Techniques of Water-Resources Investigations, Book 5; U.S. Geological Survey: Reston, VA, 1979; pp 1-626. Iowa-Cedar Riuers Basin Study; U.S. Department of Agriculture: Des Moines, IA, 1976; pp 1-250. Skow, D. M.; Holden, H. R. Iowa Agricultural Statistics; Iowa Department of Agriculture: Des Moines, IA, 1987; pp 1-111. Skow, D. M.; Holden, H. R. Iowa Agricultural Statistics; Iowa Department of Agriculture: Des Moines, IA, 1986; pp 1-130. Wintersteen, W.; Hertzler, R. In Pesticides Used in Iowa, Crop Production in 1985; Cooperative Service: Ames, IA, 198?; pp 1-18. Linsley, R. K., Jr.; Kohler, M. A.; Paulhus, J. L. H. Hydrology for Engineers; McGraw-Hill: New York, 1982; pp 1-508. Williams, G. R. In Engineering Hydraulics; Rouse, H., Ed.; Wiley: New York, 1950; pp 229-309. Interagency Advisory Committee. Hydrology Subconmittee Bulletin 17; U S . Geological Survey: Reston, VA, 1982; pp 1-28. Lara, 0. G. Water-Resour. Invest. Rep. ( U S . Geol. Surv.) 1987, NO.87-4132, 1-34. Pettyjohn, W. A.; Henning, R. Project Completion Report No. 552. Department of Geology and Mineralogy, The Ohio State University: Columbus, OH, 1979; pp 1-323. Johnston, R. H. Report of Investigation 24. Delaware Geological Survey, 1976; pp 1-56.
Environ. Sci. Technol. 1992,26, 545-552
Squillace, P. J.; Engberg, R. A. Water-Resour. Invest. Rep. (U.S. Geol. Surv.) 1988, No. 88-4060, 1-81. Singh, G.; Spencer, W. F.; Cliath, M. M.; van Genuchten, M. Th. J. Environ. Qual. 1990, 19, 520-525. Jury, W. A.; Focht, D. D.; Farmer, W. J. J. Environ. Qual. 1987, 16,422-428. Thurman, E. M. Organic Geochemistry of Natural Waters; Martinus-Nijhoff Dordrecht, The Netherlands, 1985; pp 1-497. Wu, T. L. J . Environ. Qual. 1980, 9, 459-465. Buttle, J. M. J . Environ. Qual. 1990, 19, 531-538. Wauchupe, R. D.; Myers, R. S. J . Environ. Qual. 1985,14, 132-136. Johnson, H. P.; Baker, J. L. Technical Information Service
Report P B 84-177419; Environmental Research Laboratory Office of Research and Development, U.S. Environmental Protection Agency, Athens, GA, 1980; pp 1-84. (33) Wauchupe, R. D.; Myers, R. S. J. Environ. Qual. 1985,14, 132-136.
Received for review June 12,1991. Revised manuscript received October 17, 1991. Accepted October 30,1991. The University of Iowa Hygienic Laboratory provided the herbicide analysis for this research project. Funding was provided by the Toxic Substances Hydrology Program of the U.S. Geological Survey. The use of brand names in this paper is for identification purposes only and does not imply endorsement by the U.S. Geological Survey.
Direct Mass Spectrometric Studies of the Destruction of Hazardous Wastes. 1. Catalytic Steam Re-Forming of Chlorinated Hydrocarbons Mark R. Nlmlos” and Thomas A. Milne Chemical Conversion Research Branch, National Renewable Energy Laboratory, Golden, Colorado 8040 1
Catalytic steam re-forming behavior was studied for a number of chlorinated and nonchlorinated organic species using the direct sampling capabilities of the molecular beam mass spectrometer (MBMS). Steam re-forming is an often used chemical process that can convert hydrocarbons into largely CO and H2. These screening studies investigated the possibility of using a catalytic steam reforming environment to destroy hazardous waste by measuring destruction efficiencies and qualitatively comparing product slates with straight pyrolysis. For the steam re-forming, a rhodium catalyst on a reticulated alumina support was used and these observations were made. (1) Catalytic steam re-forming significantly enhanced the destruction of the compounds studied compared to destruction by pyrolysis. (2) Products of incomplete destruction are also significantly reduced, even for those species that are thermally stable. (3) Aromatics and acetonitrile are more difficult to steam re-form than aliphatic species. (4)A wide variety of species are susceptible to steam re-forming (including chlorine-containing species and acetonitrile) without any indication of catalyst deactivation over several 10-min periods. Two exceptions identified so far are the sulfur-containing species 3methylthiophene and the phosphorus-containing species dimethyl methylphosphonate. Temperatures of the Rh catalyst must be maintained at 800 OC or above to keep the catalyst active, a t 50% steam in helium, for many compounds.
I . Introduction, Combustion of hazardous wastes is a widely practiced technology which, by most accounts, is effective in achieving the currently mandated four-nines (99.99%) destruction of principal organic hazardous constituents (POHCs) (1-5). There remain, however, substantial concerns about incineration, particularly with regard t o products of incomplete combustion (PIC). The nature of these PIC, and the circumstance of their formation, are the subject of active study in the laboratory and the field (6-10). Due to lack of continuous PIC and POHC emission monitors for field use, hazardous waste incineration is the cause of considerable environmental and public concern (as is the case with all waste incineration) (1). In addition, the rather severe combustion conditions imposed involve 0013-936X/92/0926-0545$03.00/0
1617 Cole Boulevard,
high combustion temperatures, long combustion times, and often the use of substantial quantities of blended hydrocarbon fuels. These requirements can lead to NO, formation and a substantial energy penalty. Finally, the very process of gas-phase combustion contains the reaction conditions for PIC formation under less than perfect mixing and combustor conditions. As a result of these circumstances, there is a vigorous search for alternative destruction methods ranging from pyrolysis to catalytically enhanced combustion (11-18). Among these methods is a solar-assisted approach that involves catalytic steam re-forming of a range of chemical species, with the engineering configuration involving direct irradiation of the catalyst surface with highly focused solar energy (19,20). This approach developed as an outgrowth of the C 0 2 re-forming of CH, for solar energy storage and transport, where Rh showed superior resistance to coking (19). The re-forming is carried out by having relatively cold inlet gas make contact with dispersed rhodium (21) on a reticulated, open structure of alumina (22). The steam re-forming reaction has been studied using l,l,l-trichloroethane as a model waste compound (23),where the steam re-forming stoichiometry was C,H,Cl, + 2Hz0 2CO + 3HC1+ 2H2 (1) In addition to the steam re-forming reaction, the water-gas shift reaction CO H 2 0 C 0 2 H2 (2) also occurred to produce COz. It is envisaged that few thermal gas-phase reactions will occur before the material contacts the radiantly heated surface. Direct solar irradiation of the catalyst is postulated to be beneficial, whether or not any photoenhanced heterogeneous catalytic effects are present (19,20). This is because the heat can be deposited directly on the site of the usually endothermic re-forming reactions, minimizing heat-transfer limitations. It is also possible that the absence of oxidative, gas-phase, radical-driven destruction reactions will minimize PIC formation [more generally, products of incomplete reaction (PIR)]. This system could be more forgiving of upsets, achieve high destruction efficiency (DE) at lower temperature, avoid NO, formation, and require no auxiliary fuel or hydrocarbon source of hydrogen for highly chlorinated wastes, vis-&-visincineration.
0 1992 American Chemical Society
+
-
+
Environ. Sci. Technol., Vol. 26, No. 3, 1992 545
