Environ. Sci. Technol. 1904, 18, 207-212
Inventories of 23s9240Pu 9 241Am,137Cs,and 'OCo in Columbia River Sediments from Hanford to the Columbia River Estuary Thomas M. Beasley' and C. David Jennlngs
College of Oceanography, Oregon State University, Cowallis, Oregon 97331
rn The inventories of 239*240Pu, 241Am,137Cs,and @'Coin sediments of the lower Columbia River and its estuary have been estimated from the measured activities of these radionuclides in 50 cores raised in 1977-1978. Cobalt-60 activities are attributed to the operation of the now inoperative plutonium production reactors located on the Hanford Reservation in Washington State; the ma'ority of the 2399240Pu and 137Csactivities and all of the d41Am activity are derived from global fallout. At the time of sampling, the operation of the Trojan nuclear power plant near St. Helene, OR, had not introduced detectable amounts of these radionuclides to the sandy sediments downstream from the plant location. Despite the substantial, past addition of artificial radioactivity to the river from the Hanford reactors, present inventories of artificial radioactivity are small. Erosional processes within the river's drainage basin remove negligible quantities of fallout radionuclides to the ocean.
Introduction In the period 1944-1970, large amounts of artificial radioactivity were introduced into the lower Columbia River from the Hanford Reservation (I). The source of this radioactivity was the operation of eight, single-pass plutonium production reactors which used treated river water for cooling. During this same period, above-ground testing of nuclear devices introduced a suite of radionuclides to the environment, several of which are still measurable due to the magnitude of their radioactive half-lives and the high yield of the fission chains leading to their production. Notable among these are IwCs (T1/2 = 30.2 years), BOSr (T1/2 = 28.8 years), several isotopes of plutonium, i.e., B ~ (T1/, u = 24131 years), 240Pu(T1/2 = 6570 years), and %lPu (T1/2 = 14.4 years), and 241Am = 433 years), a daughter product in the decay of 241Pu(2). For natural bodies of water that have received additions of these radionuclides via nuclear operations and global fallout, it is oftentimes difficult to quantify the contribution from each source. In previous reports ( 3 , 4 ) we have estimated the contribution of reactor Pu to the sedimentary inventory of this radioelement and the amounts of Pu and Am exported by the river to the northeast Pacific Ocean annually in the period mid-1950 through 1978. In this report, we summarize the results of radiometric measurements of m W u , 241Am,13'Cs, and @'Coin 50 cores taken from the lower Columbia River which are used to estimate the cumulative sedimentary inventory of these radionuclides downstream from the Hanford Reservation as of 1977-1978. The sources of the radionuclides are discussed (fallout vs. reactor derived), and an assessment is made of the environmental mobility of the radionuclides within the river's drainage basin. Comparison of the Columbia River data with those reported for other river systems lends additional support to the contention that for the transuranic radionuclides and 137Cs,erosional processes do not significantly alter drainage basin inventories of these radionuclides. Experimental Section Sediment samples were raised by means of a weighted 0013-936X/84/09 18-0207$01.50/0
gravity corer from the confluence of the Snake and Columbia Rivers to the Columbia River estuary (Figure 1). In all, some 57 cores were collected, and 50 of these have been analyzed and are discussed here. Upon retrieval, the sediment was extruded, sectioned (1-2-cm thickness), and 0.5 cm trimmed from the outer rim to remove downtrained material; each section was subsequently weighed, dried, homogenized, and stored in polyethylene jars prior to analysis. Generally, each third section over the core length was analyzed to establish radionuclide profiles. In the case of 23gt240P~ (and 241Am),10-20-g aliquots of the homogenized sediment were totally dissolved in a mixture of HN03-HF following the addition of calibrated amounts of "2Pu and 243Amto trace recovery of 239~240Pu and ulAm through chemical processing. Following purification, the P u and Am isolates were electrodeposited onto stainless steel disks (5) and the activities measured by a-spectrometry techniques using surface barrier diodes (a-spectrometric techniques do not distinguish between 239Puand 240Puowing to the nearly identical energies of the a particles emitted by these isotopes). 137Csand 6oCoprofiles were determined by measuring 100-g aliquots of the samples on a Ge(Li) y-ray spectrometer, a nondestructive technique which permits the determination of both isotopes simultaneously. In many of our cores, the absolute activities of the y-emitting radionuclides were too low to permit adequate quantitation by this technique. We therefore prepared composite sample for all cores by combining 5-g aliquots from each of the horizons used to establish the Pu profiles. From each of these composites, 137Csand 6oCowere determined by radiochemical purification and low-background, @-countingtechniques; m*240pu analyses were then also performed on the composite samples. The increased sensitivity of the @-countingtechniques permitted us to detect as little as 0.1 dpm/g dry weight of the 137Csand @'Coin 20-g samples, permitting positive measurements of these radionuclides in all but one of the composite samples discussed below. To determine the suitability of the composites for determining 137Csand @'Coactivities, we compared the average 2391240Pu activities over the entire core length and the 23g~240P~ activities measured in the composite samples; in all cases, agreement was excellent (f5% a t one standard deviation). After the 239*240Pu activity profiles within the cores were established, P u inventories were calculated by summing the product of the mean 239p240Pu activity (dpm/g dry weight) and the mean bulk density (g/cm3) over each 2.5 cm of core length to yield a total activity per unit area (dpm/cm2). The Pu inventories were then used to determine the inventories of 241Am,137Cs,and 6oCoin the following ways: (1)The 241Am/239*240P~ activity ratio determined in seven cores (after correction for the small amount of 239Puof Hanford origin (3)) averaged 0.29, a value we believe closely represented cumulative falloutderived activity ratios in the time period of our sampling (6). We have therefore calculated 241Aminventories from the product of this ratio and the 239*240Pu inventories. (2) 13'Cs and 6oCo inventories were calculated from their measured activities in the composite samples. Knowing
0 1984 Amerlcan Chemical Soclety
Envlron. Sci. Technol., Vol. 18, No. 3, 1984 207
I ( I /
OREGON O r - - r - r T O
Kilometars
Figure 1. Sampling locations in the lower Columbla River and Its estuary. Within McNary Reservoir (A), coring transects were required between the Oregon and Washington shores due to differing sedimentary regimes. Below McNary Dam, transects were not required.
the 2399240Pu inventories for each core and the 239J40Pu activities in the composite samples, inventories for the y-emitting radionuclides were, for example in the case of 13'Cs, calculated by using the expression
where (I) refers to the integrated inventories of the radionuclide (dpm/cm2) and (C) tQ their measured activities (dpm/g dry weight) in the composite samples. This calculation assumes that the depth profiles for all radionuclides are comparable, i.e., that either the total inventory or the same fractional amount of the inventory of each radionuclide was contained in the core analyzed; our experience has shown that this is indeed the case. Figure 2 shows the depth distribution of 2397240Pu, I3'Cs, and 6oCo in a typical core from McNary Reservoir (M-ll), confirming the strong correspondence in the activity distributions for all three radionuclides. Counting errors associated with the measurement of all radionuclides were 2154 31 651 322 23 80
>8.1 0.08 3.9 1.5 0.08 0.8
0 1176 1765 >2154e 16 747 230 29 51
E (8) F (9) a Areal activities measured as disintegrations per minute per centimeter squared have been converted to millicurie per kilometer squared for comparison with other studies cited in the text and references. Isotopic Pu analyses have shown only fallout Pu in these sediments; except for lower river sediments, only fallout Pu, Am, and Cs isotopes have accumulated in the drainage basin. See text for discussion of McNary Reservoir. Number in parentheses indicates total number of cores analyzed in each area. e In cores where measurable activities remained at depth, all core profiles showed that the large subsurface maximum in z397240Pu had been retained corresponding to the maximum fallout delivery period 1961-1963 (3). Some 75% of the total megatonnage fired in the entire period of above-ground testing occurred in this period. Assuming fallout activities are proportional to megatonnage detonated, the inventories shown here would be low by 25%. As this portion of the reservoir has 50% of the inventory of all radionuclides, increasing the inventories by 25% changes the total budget by only 12%. Error associated with inventories is t 10% (see text).
-
-
Results For convenience in presenting our estimates of inventories, we have partitioned the lower Columbia River into
Table 11. Total Radionuclide Inventories in Lower Columbia River Sediments river reach
total area, km2
Cia
241Am I3'Cs 6oC~ 172 197 2.3 0.67 6 3 0.015 0.05 28 0.15 24 0.5 18 0.12 26 0.4 9 0.03 7 0.09 23 15 0.23 F 0.8 total 4.1 1.2 254 270 a 1 Ci = 2.2 X 10l2dpm. Uncertainties associated with the individual inventory estimates are -i. 10% (see text).
A B C D E
It fi
2 3 9 9 2 4 0 P ~
116 192 37 80 309 295 1029
six reaches (Figure 1). Set out in Table I are the areal radionuclide inventories of 239*240Pu, %lAm,137Cs,and %o within each area as of 1977-1978. By far the greatest areal variation in our measured inventories occurred in McNary Reservoir, the first site of fine sediment accumulation below the confluence of the Columbia and Snake Rivers. As has been observed in other studies of this area (9), radionuclide activities were greatest in those sediments nearest the Oregon shore. In the lower reservoir, gradienta as large as 10 were observed in across-river transects between Oregon and Washington. Therefore, in establishing a budget for McNary Reservoir, we partitioned the reservoir into three equal segments, Washington/midreservoir/Oregon and averaged the inventories of cores from transects that fell into each of these segments. Below McNary Reservoir, sediment textures varied much less, and transects across the river were not made. Rather, individual cores from various sites were taken to be representative of the sediments in that area. For one reach of the river (B), we were unable to raise any cores despite repeated attempts. Completion of John Day Dam occurred in April, 1968,and was the last dam constructed in the lower river (Figure 1). Consequently, sediment accumulation has not yet occurred to any appreciable extent. The small quantities of sediment that were retrieved bore textural qualities (coarse grained) similar to those just below Bonneville Dam. We have therefore assigned, quite arbitrarily, radionuclide inventories for this river reach identical with those observed in the first core downstream from Bonneville Dam. In doing so, we emphasize that the estimate for this area must be considered a maximum. Below Bonneville Dam to the head of the estuary, all cores contained coarse-grained sand whose radionuclide content was uniformly low. The Trojan Nuclear Power Plant is located near river kilometer 140 and had been in commercial power production for some 2 years at the time of our sampling. Radionuclide inventories above and below the plant site were virtually identical, arguing for no consequential addition of either 23g,240Pu, 137Cs,or to the sediments from plant operation. Within the Columbia River estuary, fine sediment does not occur to any appreciable extent. An "average" estuary sediment is composed of about 1% gravel, 84% sand, 13% silt, and 2% clay (IO). Only in selected areas amounting to about 10% of the estuary's area do silty sediments accumulate (IO). However, in these areas, the radionuclide content of the sediments was more than an order of magnitude greater than those of the average sediments described above. Therefore, we assigned radionuclide inventories represented in fine-grained material to 10% of the estuarine area (30km2)with the remaining 90% of the area (265km2) having inventories represented by the av-
Flgure 2. Depth profiles of 239,240Pu,13'Cs, and 'OCo activities in a typical core raised from McNary Reservoir. Estimated mean sedimentation rate vs. bedrock at this site is 1.4 cmlyear. Sampling date, Aug 1977.
erage inventories of the "sand" cores. Such a partitioning resulted in half of the total radionuclide inventory residing in areas of fine sediment accumulation. In this respect, the areal inventory variation in the estuary most closely resembled those seen in McNary Reservoir than elsewhere in the river. Table I1 shows the total number of curies (Ci) of each radionuclide in the different reaches as calculated from the data of Table I and the areas assigned to the different river reaches. We have already shown that between 20 and 25% of the 23g3"P~ in McNary Reservoir sediments is reactor derived (3). By use of the higher percentage figure, about 1.7 Ci of fallout 239,240Pu has accumulated in the sediments of this reservoir. By contrast, fine-grained sediments in the estuary have been shown to contain about 3.5% reactorderived 2393240Pu ( 4 ) , about a 7-fold decrease from those observed in McNary Reservoir. This decrease must result from the importation of fallout-derived 239p240Pu to the Columbia River via the various tributaries feeding the river. Largest of these, from the standpoint of sediment discharge in the lower river, is the Willamette River (3.6 x io9 g/year (11)). Several lines of evidence lead us to conclude that the '%o we now measure in Columbia River sediments must have come largely from the activation of cobalt-containing materials in the plutonium production reactors. First, Ice Harbor Dam was completed in late Nov, 1961. Plutonium measurements in the core from this site indicate a subsurface maximum at a depth corresponding to maximum fallout delivery in mid-1963 (3).Had 6oCobeen a significant component of fallout debris during this time, its detection should have been possible in this core, yet in our composite sample, no activity above background was observed. The intervening time period between 1963 and our sampling (-15 years) would have led to substantial reduction of any e°Co activity by radioactive decay (TI 5.25 years). Therefore, we do not wish to imply that 6 0Co was never present in the sediments as a result of fallout but simply that the remaining contribution of fallout-derived 6oCoto Columbia River sediments is unimportant relative to the reactor contribution. This conclusion would seem to be contradicted by the close correspondence in the depth distribution of @Cowith those of 23g*240Pu and 13'Cs, both of which have falloutderived components (Figure 2). The two subsurface maxima correspond quite closely to the time periods 1955 and 1963,during which atmospheric testing was most intense (in terms of total megatons detonated, the U.S. tests were the most significant in the 1950s while tests conducted by the U.S.S.R. dominated above-ground testing Environ. Sci. Technol., Vol. 18, No. 3, 1984
209
Table 111. Activity Levels of Reactor-Produced Radionuclides in Columbia River Water Downstream from Hanford pCi/L year
slCr
1959 1960 1961 1962 1963 1964 1965 1966 1967
4198 5404 5774 4285’ 6736 5722 4043 2671 2375
32P 65Zn 154 210 264 176 194 86 87 93 85
206 298 341 224 308 144 161 135 140
reference HW-64371(1960)“ HW-68435(1961) HW-71999(1962) HW-76526(1963) HW-80991(1964) BNWL-90 (1965) BNWL-316 (1966) BNWL-439 (1967) BNWL-983 (1969)
a All reports in this series are entitled “Evaluation of Radiological Conditions in the Vicinity of Hanford for (year)”. HW = Hanford Works; BNWL = Battelle Northwest Laboratories. We can find no explanation in either the 1963 or 1964 reports explaining the decrease in activities seen in 1962.
during the early 1960s (12)). The explanation for this rests, we believe, with the fact that the Hanford reactors were a principal source of Pu for both nuclear device testing and for nuclear weapons stockpiling. In those periods of greatest testing, and weapons production, it is easy to envision that maximum operation of the reactors was required. During such periods, production and release of artificial radioactivity to the river would have been greatest. Intervening periods would have entailed less intense P u production schedules. As a reactor-produced radionuclide, we view the e°Co depth distributions in Figure 2 as being directly related to the operating history of the reactors. This interpretation is supported by reference to the mean, annual activities of short-lived radionuclides of reactor origin measured in Columbia River water downstream from Hanford a t a time when all eight production reactors were in operation (Table 111). The general trend of increasing activities in the period 1959-1963, and the decreasing activities thereafter, is consistent with the arguments presented above. Beginning in 1965, certain ones of the reactors were shut down, which accounts for the decreases observed after that date. The sources of 137Csin the river sediments are less easily quantified. The plutonium production reactors did contribute 137Csto the sedimentary inventories, however, on the basis of the following considerations. When the 239p240Pu sediment inventory of fallout origin, the total inventory of 137Csin the sediments, and the cumulative 137Cs/239v240Pu activity ratio (henceforth shown only as Cs/Pu) in global fallout through 1977 are known, the 137Cs contributed by the reactors to river sediments can be calculated from the expression 137cs(R)= 137cs(T)- (CS/PU)(F) x Pu(J?)
(2)
where 137cs(R) is the total reactor contribution to the sedimentary inventory, 137Cs(T)constitutes the total 137C~ sediment inventory, (CS/PU),, is the activity ratio of these isotopes from cumulative fallout through 1977, and PUQ) represents the 2393240Pu fallout sediment inventory (13). When 137cs(T) and PqF)have been determined, it remains then to choose a value for (Cs/Pu), b estimate the reactor contribution of 137Csby using eq 2. Estimates of the cumulative fallout ratio a t New York City are calculated to have been 67 in late 1975 (14), while soil measurements in the northern temperate zone as late as 1979 gave Cs/Pu ratios near 53 (13). Incremental increases in 137Csfallout over the period 1971-1979 are believed to have nearly 210
Envlron. Scl. Technol., Vol. 18, No. 3, 1984
balanced 137Csdecay in this period (13),and therefore, only slight changes in the Cs/Pu activity ratio should have occurred during the years of our sampling. Assuming that the Cs/Pu ratio in fallout was as high as 67, substitution of values from Table I1 for McNary Reservoir sediments into eq 2 (using a 2391240Pu fallout inventory of 1.7 Ci) gives a calculated reactor 13Ts contribution of 58 Ci to the total 172 Ci of 137Cswithin the reservoir sediments. As we are unaware of any measurements that would assign higher Cs/Pu ratios for cumulative fallout at the time of our sampling, we conclude that a t least this amount (58 Ci) of 137Csmust be reactor derived. The same calculation using a Cs/Pu ratio of 53 yields a reactor-derived 13’Cs contribution of 82 Ci. The calculation shown here presupposes that no fractionation occurs between 137Csand 23g7240Pu in the river system, i.e., that cumulative fallout ratios as measured in cores from undisturbed soil are maintained in erosional and runoff processes. Such a fractionation has been inferred for these processes in the Hudson River drainage basin (14). If correct, global fallout ratios of Cs/Pu in the Columbia would be lowered due to an enhanced mobility of 137Csover 2393240Pu.Our one core from Ice Harbor Reservoir shows a Cs/Pu activity ratio of 44, suggesting such a fractionation might indeed be occurring. For the present, however, we lean toward an estimate of between 58 and 82 Ci as the amount of 137Csof reactor origin in McNary Reservoir sediments. At this time, we have not made measurements of the isotopic composition of Pu isolated from Sediments other than those in McNary Reservoir and the estuary. We are not able, therefore, to estimate the amount of 13Tsin the other river reaches that has come from reactor operations. However, there must be a general reduction in the percentage of both 239*240Pu and 13Ts in sediments downstream from Hanford due to their importation by soil erosion and runoff. ’In the estuary itself, 137Csinventories are reduced with respect to 2399240Pu inventories (Cs/Pu 29; Table 11). It is generally thought that this results, principally, from the desorption of 137Csfrom the sedimentary material under the influence of saline waters (ref 14 and 15, and references therein); by contrast, there is a growing body of evidence to suggest that little, if any, postdepositional migration of P u occurs in either freshwater (ref 14, and references therein) or marine sediments (15-17). We reemphasize that the inventories of Table I, as computed by using eq 1, used the measured Cs/Pu ratios of each composite sample, and therefore, even though Cs/Pu ratios are different between the upper river and the estuary, the total inventory of 137Csin the estuary cores can be correctly calculated. What is compromised by this altered estuarine ratio is our ability to estimate the percentage of reactor-derived 137Csin the estuary.
-
Discussion All of the radionuclides discussed here have been shown to form strong associations with particulate matter in freshwater systems (18-21). Measurements in the Columbia River of the fractional adsorption of Pu and Am ( 4 ) and @%30 (20,21)to suspended particulate matter have indicated that upon filtration through 0.3-pm Millipore filters, 80-90% of the transuranic radionuclides and between 80 and 95% of the 6oCoare retained on the filters. Our own determination of the distribution coefficient, Kd (Kd = (activity/gram of sediment)/ (activity/gram of water)) (22), for 137Csusing filtered (0.3-pm) Columbia River water and McNary Reservoir sediments (sediments from two different locations with suspended loads of 60 mg/L) give values near 2 x lo3, which shows the strong
adsorption characteristics of these sediments for this isotope. This is entirely consistent with the findings of Gillam et al. (23),who demonstrated that soils containing montmorillonite as a major or intermediate species of abundance in their clay assemblages exhibit Kd)s in the range 103-104 for 137Cs;montmorillonite is the major (>50%) clay species in the clay assemblages of lower Columbia River sediments (24,25). Indeed, montmorillonite/chlorite ratios have been used as a tracer of sediment discharged from the Columbia River to study sediment transport on the continental shelf off Washington (26). The total area of the Columbia’s drainage basin is -6.7 x lo6 km2 (24). Extensive dam construction on both the Snake and Columbia Rivers serve principally to slow downstream sediment transport rather than to interdict it. Fallout input of 239*%F’uat 45O N latitude through 1976 was approximately 2.5 mCi/km2 (27),giving a total inventory of 1700 Ci in the drainage basin. The 241Am inventory associated with this quantity of 239p240Pu would be near 490 Ci. From other studies (4),we have estimated that since the mid-1950s through 1979, the Columbia River exported a maximum of 8 Ci of 239*240Pu and slightly more than 2 Ci of 241Amto the northeast Pacific Ocean, principally in association with the sediment carried by the river. Given this small export, the transuranic inventories of the drainage basin, can, as a first approximation, be considered as constant. Thus, the mean removal rate for P u approximates 0.3 Ci/year and that for Am, 0.08 Ci/ year. The calculated “response time” (28) for transuranic removal from the landscape by erosional processes is thus near 6000years (response time = inventory/removal rate). Assuming a Cs/Pu ratio of 53 (13),the calculated fallout inventory of 137Csin the river’s drainage basin in 1977 was approximately 9.0 X lo4Ci. We estimate that as much as 500 Ci of 137Cscould have been exported by the river to the ocean (200 Ci with sediment discharge (15) and perhaps as much as 300 Ci in the dissolved phase (14) plus desorption from estuarine sediments), giving a mean removal rate of 22 Ci/year. By use of these estimates, the calculated response time for 137Cs removal from the drainage basin by erosion is near 4000 years. Given the radioactive half-lives of 2399240Pu, 241Am,and 137Cs,the majority of the 137Cswill decay within the drainage basin before transport to the river and then to the ocean, while 2399240Pu and 241Amwill continue to be eroded from the landscape for centuries. The fate of the radionuclides already in the river sediments (as distinct from the landscape comprising the drainage basin) will be largely dictated by the fate of the sediments themselves. The majority of the radionuclide inventories reside behind dam sites (Table 11). Given continued sediment accumulation and no new anthropurgic additions, 6oCoinventories would be reduced to l/lm of their 1977-1978 values in just over 50 years. The same fractional reduction in 137Csinventories would require approximately 300 years; these times become 4330 and >240000 years, respectively, for 241Amand 239Pu.As the longevity of the dams themselves is not known (with any certainty) over time periods exceeding several centuries, the ultimate fate of the long-lived radionuclides cannot be addressed. Our findings concerning the mobility of the transuranic radionuclides and 137Cswithin this large drainage basin are consistent with those found for the drainage basins of the Great Miami River (29) (area = 1.4 X lo3 km2), the Savannah River Basin (30)(area = 2.7 X lo4km2),and the Hudson River (14) (area = 3.5 X 104km2). In none of these has substantial movement from the landscape occurred.
Finally, and unrelated to the question of environmental mobility, it is instructive to note that for a river which, in the mid-l960s, received as much as 300000 Ci/year of artificial radioactivity from the operation of the reactors at Hanford (9),the residual radioactivity we measure in the sediments some 13 years later is vanishingly small. Radioactive decay of short- and moderate-lived radionuclides, sediment export, and the large flow of the river (2.3 X 1014L annually (4))have combined to reduce activities to levels that are no longer easily determined. In fact, recent measurements of both natural and artificial radioactivity in sediments raised from McNary Reservoir (31) have shown that the natural radioactivity of the sediments (containing isotopes of K, Th, U, and Ra) exceeds that of the artificial radioactivity by nearly an order of magnitude.
Acknowledgments 239w~
N
We acknowledge the help of many colleagues in this work. Sample collection and preparation (N. Farrow, D. Higley, J. Morgan, and H. Batchelder) and the radiochemical analyses (L. Ball, B. Blakesley, J. Andrews, 111, and H. Batchelder) were both time consuming and demanding. R. Carpenter, N. Cutshall, and P. Krey made valuable comments on the manuscript which greatly improved its organization and clarity. Registry No. 238Pu,15117-48-3; 240Pu,14119-33-6; 241Am, 14596-10-2; 13’Cs, 10045-97-3;6oCo,10198-40-0.
Literature Cited (1) Foster, R. F. In “The Columbia River Estuary and Adjacent Ocean Waters”; Pruter, A. T.; Alverson, D. L., Eds.; University of Washington Press: Seattle, WA, 1972; pp 3-18. (2) “Table of Isotopes”, 7th ed.; Lederer, M. D.; Shirley, V. S., Eds.; Wiley: New York, 1978. (3) Beasley, T. M.; Ball, L. A.; Andrews, J. A., 111.;Halverson, J. E. Science (Washington,D.C.) 1981,214, 913-915. (4) Beasley, T. M.; Ball, L. A.; Blakesley, B. A. Estuarine, Coast Shelf Sci. 1981, 13, 659-669. (5) Talvitie, N. A. Anal. Chem. 1972, 44, 280-283. (6) Beasley, T. M.; Ball, L. A. Nature (London) 1980, 287, 624-625. (7) Jennings, C. D.; Beasley, T. M. Tulanta 1982,29,871-873. (8) Volchok, H. L.; Feiner, M. U.S. Department of Energy, New York, 1980. Environmental Measurements Laboratorv Report EML-366, pp 1-43. Robertson, D. E.; Silker, W. B.; Langford, J. C.; Petersen, M. R.; Perkins, R. W. “Radioactive Contamination of the Marine Environment, Proceedings of a Symposium”; International Atomic Energy -. Agency: - - Vienna, Austria, 1973; pp 141-158. Hubbel, D. W.; Glenn, J. L. Geol. Suru. Prof. Pap. (US.) 1973, NO.433-L, 1-35. Karlin, R. J. Sediment. Petrol. 1980, 50 (2), 543-560. Carter, M. W.; Moghissi, A. A. Health Phys. 1977,33,55-71. Krey, P. W.; Beck, H. L. U.S. Department of Energy, New York, 1981, Environmental Measurements Laboratory Report EML-400, pp 1-45. Olsen, C. R.; Simpson, H. J.; Trier, R. M. Earth Planet. Sci. Lett. 1981,55, 377-392. Beasley, T. M.; Carpenter, R.; Jennings, C. D. Geochim. Cosmochim. Acta 1982,46, 1931-1946. Carpenter, R.; Beasley, T. M. Geochim. Cosmochim. Acta 1981,45, 1917-1930. Santschi, P. H.; Li, Y. H.; Adler, D. M.; Amdurer, M.; Bell, J.; Nyffeler, U. P. Geochim. Cosmochim. Acta 1983, 47, 201-210. Watkrs, R. L.; Edgington, D. N.; Hakonson, T. E.; Hanson, W. C.; Smith, M. H.; Whicker, R. W.; Wildung, R. E. In “Transuranic Elements in the Environment”; Hanson, W. C., Ed.; Technical Information Center: Springfield, VA, 1980; pp 1-45. Environ. Sci. Technol., Vol. 18, No. 3, 1984
211
Environ. Sci. Technol. 1984, 18, 212-216
Edginpton, D. N.; Alberta, J. J.; Walhgren,M. A.; Karttunen, J. 0.; Reeve, C. A. “Transuranium Nuclides in the Environment, Proceedings of a Symposium”; International Atomic Energy Agency: Vienna, Austria, 1976; pp 493-516. Nelson, J. L.; Perkins, R. W.; Nielsen, J. M.; Haushild, W. L. “Disposal of Radioactive Wastes into Seas, Oceans, and SurfaceWaters, Proceedings of a Sympmium”;International Atomic Energy Agency: Vienna, Austria, 1966; pp 139-161. Perkins, R. W.; Nelson, J. L.; Haushild, W. L. Limnol. Oceanogr. 1966, 11, 235-248.
Beasley, T. M., Oregon State University, Corvallis, OR, unpublished data, 1983. Gillham, R. W.; Cherry, J. A.; Lindsay, L. E. Health Phys.
(27) U.S. Energy Research and Development Administration, New York, 1977, Health and Safety Laboratory Quarterly Report HASL-329, pp 1-401. (28) Drever, J. I. “The Geochemistry of Natural Waters”; Prentice-Hall: Englewood Cliffs, NJ, 1982; Chapter 1. (29) Sprugel, D. G.; Bartelt, G. E. J . Environ. Qual. 1978, 7 (2), 175-177. (30) Hayes, D. W.; Horton, J. H. In “TransuranicElements in
the Environment”;Hanson, W. C., Ed.; Technical Information Center: Springfield, VA, 1980; pp 602-611. (31) National Bureau of Standards, Standard Reference Material 4350B, River Sediment, 1982.
1980,39,637-649.
Knebel, H. J.; Kelley, J. C.; Whetten, J. T. J . Sediment. Petrol. 1968, 38, 600-611.
Whetten, J. T.; Kelley, J. C.; Hanson, L. G. J . Sediment. Petrol. 1969, 39, 1149-1166. Baker, E. T. Geol. SOC.Am. Bull. 1976,87, 625-632.
Received for review February 18, 1983. Revised manuscript received August 11, 1983. Accepted September 12, 1983. Our research is funded by the Environmental Research Division, Office of Health and Environmental Research, US.Department of Energy.
Stoichiometry of the Reaction between Chlorite Ion and Hypochlorous Acid at pH 5 Tsung-fel Tang and Gilbert Gordon” Department of Chemistry, Miami University, Oxford, Ohio 45056
The availability of an improved analytical method allowed a preliminary study of the stoichiometry of the reaction between chlorite ion and hypochlorous acid at pH 5. The stoichiometry varies with the initial concentration ratio of the reactants, suggesting parallel pathways for the reaction mechanisms. The yield of chlorine dioxide is shown to vary with the initial concentration ratio of hypochlorous acid to chlorite ion as well as the contact time between reactants. The formation of molecular oxygen is postulated as a byproduct.
Introduction The major method used by water treatment plants for the generation of chlorine dioxide is mixing aqueous solutions of chlorine and sodium chlorite (1-4).Granstrom and Lee (2) surveyed 56 water treatment plants using chlorine dioxide. The ratio of chlorine to sodium chlorite used by these water treatment plants for the generation of chlorine dioxide varied from 1 0 1 to 1:l. The survey also indicated only 25 plants measured the pH of the solution and that the effluent pH of the chlorine dioxide generator varied from 1.2 to 10.4. The present experimental results demonstrate that chlorate ion is generated at pH 5 and that side reactions also consume chlorine dioxide and form oxygen as well as chlorate ion. The stoichiometric details of this reaction above pH 5 are not clearly understood since other products in addition to chlorine dioxide may be formed (2). In order to study the overall reaction, it is necessary to determine quantitatively the various oxychlorine species, namely, hypochlorous acid (hypochlorite ion), chlorite ion, chlorine dioxide, and chlorate ion. The analytical determination of these oxychlorine species is complicated due to their similar chemical and physical properties. Chlorine hydrolyzes in aqueous solution at pH 5 to form hypochlorous acid. A preliminary kinetic study of the reaction between chlorite ion and hypochlorous acid was carried out by Granstrom et al. (5). However, they re212
Envlron. Sci. Technol., Vol. 18, No. 3, 1984
ported many experimental difficulties were involved. The analytical method they used for the determination of chlorine dioxide, chlorite ion, and hypochlorous acid was the spectrophotometric method, but due to the overlap of the various spectra in the UV region, the computation for the concentrations of the oxychlorine species necessitated the solution of three simultaneous equations. A slight error in the value of the molar absorptivity or of the observed absorbance is magnified considerably in the final values obtained for the concentrations of hypochlorous acid, chlorite ion, and chlorine dioxide. Another problem associated with the analytical method reported by Granstrom et al. (5) is the determination of total chlorine in the solution. The total chlorine was determined by passage of the sample through a Jones reductor, and the resulting chloride ion formed was titrated potentiometrically with silver nitrate. However, the high volatility of chlorine dioxide increased the difficulty of sample handling, and some chlorine dioxide invariably was lost to the atmosphere (5). Thus, an error of nearly 5% was reported for the total chlorine determination. In the present study, the stoichiometry of the reaction between chlorite ion and hypochlorous acid a t pH 5 was determined under various reactant ratios. The experimental difficulties described by Granstrom were minimized by using an improved analytical procedure (6). The procedure involves the combination of two analytical methods, namely, the ion selective electrode method (6) and the iodometric method (1, 7). The iodometric method is used to determine the oxychlorine species in terms of redox properties, and the ion selective electrode method is used to determine the chloride ion, chlorate ion, and the total chlorine by the halogen content. Thus, the hypochlorous acid-chlorite ion reaction can be studied more precisely by utilizing two different properties.
Experimental Section
Reagents. Reagent-grade chemicals and triply deionized distilled water were used throughout. Preparation and
0013-936X/84/0918-0212$01.50/0
0 1984 American Chemical Society