Envlron. Scl. Technol. 1992, 26, 1655-1658
(12) Munder, A.; Buchert, H.; Niemczyk, R.; Swerev, M.; Ballschmiter, K. Fresenius 2. Anal. Chem. 1987, 328, 639-643. (13) NATO/Committee on the Challenges of Modern Society: International Toxicity Equivalency Factor (I-TEF)Method of Risk Assessment for Complex Mixtures of Dioxins and Related Compounds. Report No. 176 and Report No. 178; EPA No. 68-02-4254; US. Environmental Protection Agency (EPA), Office of Research and Development, and V e n a Inc., U.S. Government Printing Office: Washington, DC, 1988. (14) Swerev, M.; Ballschmiter, K.; Fresenius, Z. Anal. Chem. 1987,328, 125-127. (15) Ballschmiter, K.; Braunmiller, I.; Niemczyk, R.; Swerev, M. Chemosphere 1988,17,995-1005. (16) Swerev, M.; Ballschmiter, K. Chemosphere 1989, 18, 609-616. (17) Rappe, C.; Bergkvist, P. A.; Kjeller, L. 0. Chemosphere 1989,18, 651-658. (18) Edgerton, S. A.; Czuczwa, J. M.; Rench, J. D.; Hodanbosi, R. F.; Koval, P. J. Chemosphere 1989, 18, 1713-1730. (19) Czuczwa, J. M.; Hites, R. A. Environ. Sci. Technol. 1984, 18,444-450. (20) Hashimoh, S.;Wakimoho, T.; Tatsukawa, R. Chemosphere 1990,21, 825-836.
Literature Cited (1) Bumb, R. R.; Crummet, W. B.; Cutie, S. S.; Gledhill, J. R.; Hummel, R. H.; Kagel, R. 0.;Lamparski, L. L.; Luoma, E. V.; Miller, D. L.; Nestrick, T. J.; et al. Science 1980, 210, 385-390. (2) Nestrick, T. J.; Lamparski, L. L. Anal. Chem. 1982, 54, 2292-2299. (3) Thoma, H.,Chemosphere, 1988, 17, 1369-1379. (4) Chin, C.; Thomas, R. S.; Lockwood, J.; Li, K.; Halman, R.; Lao, R. C. C. Chemosphere 1983,12, 607-616. (5) Ballschmiter, K.; Bacher, R.; Riehle, U.; Swerev, M. Technical Report No. 07640040. Beitrag der Automobilabgase zu der allgemeinen Umweltbelastung durch polychlorierte Dibenzodioxine (PHalDD) und Dibenzofurane (PHalDF). Forschungsbericht Umweltforschung BMFT, Bonn, 1990. (6) Swerev, M. Dr. rer. nat. Dissertation, University Ulm, 1988. (7) Harrad, S. J.; Fernandez, A. R.; Creaser, C. S.; Cox, E. A. Chemosphere 1991,23, 255-261. (8) Zoller, W.; Ballschmiter, K. Chemosphere 1986, 15, 2 129-2 132. (9) Riehle, U. Dr. rer. nat. Dissertation, University of Ulm, 1990. (10) Ballschmiter, K.; Buchert, H.; Niemczyk, R.; Munder, A.; Swerev, M. Chemosphere 1986, 15, 901-915. (11) Hagenmaier, H.; Brunner, H.; Haag, R.; Berchtold, A. Chemosphere 1986, 15, 1421-1428.
Received for review December 3, 1991. Revised manuscript received March 23, 1992. Accepted March 26, 1992.
COMMUNICATIONS Concentrating Suspended Sediment Samples by Filtration: Effect on Primary Grain-Size Distribution Ian 0 . Droppo," Bommanna G. Krlshnappan, and Edwln D. Ongley
National Water Research Institute, 867 Lakeshore Road, P.O. Box 5050, Burlington, Ontario, Canada L7R 4A6 polycarbonate plastic membrane filters).
Introduction
Suspended solids play an important role in the biological and chemical dynamics of the aquatic environment. They are also identified as important components for the transport of contaminants in rivers (1,2). Because many contaminants demonstrate a high affinity for the fine-grain fraction of sediments, the primary grain-size distribution is often of interest for sediment and contaminant transport studies. The accurate determination of suspended grainsize distributions by traditional methods is often difficult if sediment concentrations are low. Common methods of concentrating suspended sediments for analysis are evaporation and centrifugation. Evaporation is, however, time consuming and may alter the grain-size distribution by precipitation of dissolved solids. Centrifugation is a viable alternative and is recommended when chemical analysis of suspended sediment is required (3). However, if the purpose of sediment analysis is to provide information on primary grain-size distribution and not suspended solid concentration or chemistry, then filtration and resuspension of the sediment from filters may be suitable for analysis with many sediment sizing techniques. The objective of this communication is to determine whether the primary grain-size distribution is altered by the process of filtering and resuspending sediment off of two types of filters (Millipore cellulose membrane filters and Nuclepore 0013-936X/92/0926-1655$03.00/0
Materials and Methods
Sample Preparation. Bulk bottom sediment collected from Lake Erie offshore of Port Stanley and Port Burwell was used in this study. The sediment was fully dispersed, wet-sieved through a 62-pm mesh to ensure a size distribution in the silt and clay range and then freeze-dried. Organic content was 2.9% as determined by loss on ignition. Rather than produce six separate solutions with six subsamples from the dry sediment, one solution was produced and six subsamples were drawn off to allow for a more even distribution of particle sizes between subsamples. A 100-mg sample of the bulk sediment was suspended in 1L of 10% solution of sodium hexametaphosphate in distilled water. The solution was sonicated for 2 min to disperse the sediment into its primary particles and then placed on a magnetic stirrer to keep the particles in suspension. Six 10-mL aliquots were withdrawn by pipeting from the dispersed sedimentlwater solution and successively placed into six beakers until the original suspension was depleted. Sampling depth was the same for each round of six withdrawals to account for any possible segregation of size classes within the suspension. Additional distilled water was added to retrieve any particles de-
0 1992 American Chemical Society
Environ. Sci. Technol., Vol. 26, No. 8, 1992
1655
posited on the bottom of the original solution beaker and distributed (by pipeting) evenly among the six samples. Each sample was approximately 170 mL with a calculated concentration of 16.7 mg/L. A small portion (1-3 mL) of each of the six samples (representing the initial prefiltered solution) was removed (after sonication) by pipet for particle sizing with the Malvern particle size analyzer (Series 2600~).The initial subsample was then filtered through the respective filter types and the sediment immediately washed off with distilled water spray before the filter and sediment could dry. The resuspended solutions were once again sonicated and vigorously shaken prior to particle-size determination. Through triplicate measurements, the precision of the Malvern readings of median grain size was found to be within 1 pm. Filters. Two common filter types (Nuclepore and Millipore) were used to determine if there is any difference in their ability to retain sediment or to degrade during the sediment recovery process and thus affect the grain-size distribution. Nuclepore Corp. does not produce a 47mm-diameter polycarbonate plastic membrane filter with a nominal pore size of 0.45 pm (the conventional boundary between the dissolved and particulate phase). Therefore, the closest pore size (0.4 pm) was used. The filter has cylindrical pores normal (within f29') to the surface and randomly distributed. The pore size deviation is +O to -20% of the nominal pore size. The majority of sediment is retained on the surface of the filter due to its flat smooth surface (4). The Millipore filter (47-mm diameter) is a membrane fiter composed of cellulose acetate and cellulose nitrate fibers woven into a matrix with a nominal pore size of 0.45 pm. This filter is a weave of fibers with many flow channels which retains particles within its matrix of fibers as well as on the surface (5). Particle Sizing. Primary particle grain-size distributions were measured using a Malvern particle size analyzer (Series 2600~). The Malvern consists of a 3-mV laser, receiving optics assembly, and electronic circuitry interfaced with a microcomputer. Particle-size determinations are derived in less than 1 min from measurements of the near-forward Fraunhofer diffraction spectrum that is provided by a particle group randomly distributed in a sample cell mounted in the beam path between the laser source and the detector array ( 6 ) . The minimum concentration for optimum operation of the Malvern particle size analyzer depends on the size of particles and on the path length of the laser beam (7). For an average particle diameter of 10 pm and path length of 10 mm, the minimum volumetric concentration is 3.4 ppm. A more complete description of the Malvern particle size analyzer can be found in Krishnappan et al. (8).
Results and Discussion Initial primary grain-size distributions (by volume) of the six samples are plotted in Figure 1, indicating an envelope of variability for particle-size distributions between samples. A one-way analysis of variance (ANOVA; a = 0.01) reveals no significant difference between the means of the distributions for each of the six samples. Any variation between samples is likely due to random factors, including unequal particle size sampling (splitting) from the original 1-L suspension and instrumental factors. Variance in particle frequency was greatest in the larger particle-size classes and decreased with particle size. This reflects the smaller number of larger particles in the distribution resulting in increased variability in the larger size classes. Particles greater than 62 pm in the distribution may be due to flocculation occurring in the suspension 1656 Environ. Sci. Technol., Vol. 26, No. 8, 1992
Table I. Initial vs Resuspended Mean Sample Distribution Variations for Millipore Filters
size class, pm 118.4-54.9 54.4-33.7 33.7-23.7 23.7-17.7 17.7-13.6 13.6-10.5 10.5-8.2 8.2-6.4 6.4-5.0 5.0-3.9 3.9-3.0 3.0-2.4 2.4-1.9 1.9-1.5 1.5-1.2
init distrib % SD" 12.3 23.3 10.2 9.9 6.5 4.9 5.4 5.0 4.3 4.6 4.4 3.1 1.9 1.1 0.8
2.6 2.7 0.7 0.5 0.6 0.8 0.6 0.7 0.7 0.7 0.8 0.6 0.4 0.2 0.1
resusp distrib % SD 16.5 26.3 10.4 8.8 5.9 4.4 4.3 4.1 3.9 4.0 3.7 2.8 1.6 0.8 0.6
3.1 1.1 1.1 0.1 0.7 0.3 0.6 0.5 0.6 0.7 0.7 0.6 0.4 0.2 0.2
difference SD
70
+25.4 +11.4 +1.9 -11.1 -9.2 -10.2 -20.4 -18.0 -9.3 -13.0 -15.9 -9.7 -15.8 -27.3 -25.0
+16.1 -59.2 +36.4 -80.0 +14.3 -62.5 0.0 -28.6 -14.3 0.0 -12.5 0.0 0.0 0.0 +100.0
SD, standard deviation ( N = 3). Table 11. Initial vs Resuspended Mean Sample Distribution Variations for Nuclepore Filters
size class, pm 118.4-54.9 54.4-33.7 33.7-23.7 23.7-17.7 17.7-13.6 13.6-10.5 10.5-8.2 8.2-6.4 6.4-5.0 5.0-3.9 3.9-3.0 3.0-2.4 2.4-1.9 1.9-1.5 1.5-1.2
init distrib % SD" 18.6 30.6 10.2 9.0 5.4 3.7 3.7 3.2 2.9 3.1 3.1 2.1 1.2 0.8 0.8
2.8 2.7 1.6 0.5 0.8 0.8 0.5 0.6 0.5 0.4 0.4 0.3 0.2 0.1 0.1
resusp distrib % SD 21.1 31.4 10.2 8.7 4.9 3.5 3.5 3.0 2.8 2.9 2.6 1.9 1.1 0.7 0.6
1.9 0.4 0.7 0.3 0.7 0.6 0.5 0.4 0.2 0.2 0.1 0.2 0.2 0.2 0.2
difference SD
%
+11.8 +2.5 0.0 -3.3 -9.2 -5.4 -5.4 -6.2 -3.4 -6.4 -16.1 -9.5 -8.3 -12.5 -25.0
-32.1 -85.2 -56.2 -40.0 -12.5 -25.0 0.0 -33.3 -60.0 -50.0 -75.0 -33.3 0.0 +100.0 +100.0
" SD, standard deviation (N = 3). prior to analysis or the presence of elongated particles which have passed longitudinally through the sieve but whose long axis was sized by the Malvern. Distributions of the resuspended suspensions are plotted in Figures 2 and 3 for each filter type and compared to their initial sample distributions. Once again visual observations reveal variation in the resuspended distributions to be greatest in the larger size classes. Figure 2 and 3 illustrate significant overlap in the envelopes of the resuspended and initial distributions. The percent differences between the initial and resuspended percent sediment in each size class given in Tables I and 11,however, demonstrate an increase in the coarser particle ranges (23.7-118.4 pm) with a corresponding decrease in the finer fractions (1.2-23.7 pm). While the percent differences (of the percent in each range) are relatively large (up to 27%),the relative importance of these differences to the overall particle size distribution appears minimal (Figures 2 and 3). Tables I and I1 also demonstrate a large percent difference between the standard deviations of the initial and resuspended distribution size classes. The majority of these percent differences are negative (Le., resuspended standard deviations are smaller than the initial standard deviations), indicating a narrowing of the size spectrum. This may be reflective of a flocculation
I
30 25
s
a,
-
20
n>, w
15
9
10
5
0
118.4- 54.9- 33.7- 23.7- 17.7- 13.6- 10.5- 8.254.9 33.7 23.7 17.7 13.6 10.5 8.2 6.4
6.4- 5.0 - 3.93.9 3.0
5.0
3.02.4
2.4- 1.91.9 1.5
1.51.2
Size classes ( l m ) Flgure 1. Grain-size distribution variability of initial subsampies. "d
30 initial distribution
25
0 Resuspended distribution
a,
-5
20
3
15
B
ae
10
5 n U p -
118.4- 54.9- 33.7- 23.7- 17.7- 13.6- 10.5- 8.254.9 33.7 23.7 17.7 13.6 10.5 8.2 6.4
6.45.0
5.0 3.9
3.93.0
3.02.4
2.4- 1.91.9 1.5
1.5-
1.2
Size classes ( pm) Flgure 2. Grain-size distribution of initial and resuspended sediment subsampies for Nuclepore filters.
35 initial distribution
aResuspended distribution
a,
-E,
B
n>, 9
"
118.4- 54.9- 33.7- 23.7- 17.7- 13.6- 10.5- 8.254.9 33.7 23.7 17.7 13.6 10.5 8.2 6.4
6.4- 5.0 - 3.95.0 3.9 3.0
3.02.4
2.4- 1.91.9 1.5
1.5-
1.2
Size classes ( l m ) Flgure 3. Grain-size distribution of inliai and resuspended sediment subsamples for Miillpore filters.
process occurring during filtration, shifting the spectrum of the distribution to the higher size classes or to the resuspension of coagulated sediment from the filter which was not completely disaggregated by sonication. The increased percentage of particles in the larger size class (23.7-118.4 pm)would tend to support one or both of these theories. Visual observation of the resuspended suspension from the Millipore filters revealed some fibrous material present from the filter structure, suggesting filter disintegration. This may also explain the higher percent by volume of particles in the larger size classes as compared
to the initial sediment for this filter. This problem may be minimized by prewashlng the filters. Within the smaller size classes there tends to be slightly fewer particles by volume in the resuspended sediment than in the initial sediment suspension. This is indicative of some pore clogging and sediment retention by both filters. Horowitz (3) found that back-flushed filters retained greater than 40-70% of the sediment filtered. Visual observations of both filter types after washing revealed discoloration induced by sediment retention. This phenomenon was more pronounced in the Millipore filter Environ. Sci. Technoi., Voi. 26, No. 8, 1992
1657
Environ. Sci. Technol. 1992,26,1658-1662
due to its weave construction. The difference between the initial and resuspended sediment distributions for both filter types, even with the aforementioned pore clogging, filter disintegration, and sediment flocculation or coagulation, is statistically insignificant (ANOVA, a = 0.01). There is also no significant difference between the resuspended sediment distributions for the two filter types (ANOVA, cy = 0.01), indicating that the type of filter used to concentrate the sediment has a minimal influence on the resultant primary particle distribution, even with the differences in nominal pore sizes. The minimal influence of pore clogging on the grain-size distribution may be explained by the fact that there is a greater number of smaller particles than larger particles in the distribution. As it is the small sediment sizes which are most likely to be trapped by the pores of the filters, the impact of their loss to the distribution appears to be unimportant. The filtration of the 5 psi) may bind particles to the filters as well as suck particles through the filter, resulting in increased deviation from the actual primary grain-size distribution. The length of time required to complete the described technique is mostly dependent on the filtration rate, which is in turn highly dependent on the suspended sediment concentration. If a Millipore fibrous membrane filter type is used, it is recommended that the filters be prewashed to minimize filter disintegration and distribution distortion. Conclusions Often suspended primary grain-size distributions are required for sediment/contaminant transport studies. This parameter is, however, often difficult to quantify when suspended sediment concentrations are low. If this condition exists, the sediment must be concentrated for a more accurate determination of particle size. On the
basis of the results of this study, it can be concluded that concentrating predominantly fine inorganic sediments by filtering on and resuspending from both Millipore cellulose membrane (0.45 pm) and Nuclepore polycarbonate membrane (0.4 pm) type filters does not significantly alter the initial primary grain-size distributions. While some filter disintegration and pore clogging occurred, their influence on the primary grain-size distribution was minimal. Acknowledgments We thank R. Stephens for carrying out the size distribution measurements using the Malvern particle size analyzer and Dr.S. S. Rao, J. Marsalek, and M. Stone for their review of the manuscript. The comments of the three anonymous reviewers are also appreciated. Literature Cited (1) Allan, R. J. T h e Role of Particulate Matter in the Fate of
(2)
(3)
(4) (5)
(6) (7)
(8)
Contaminants in Aquatic Ecosystems; Inland Waters Directorate, Scientific Series No. 142; National Water Research Institute, Canada Centre for Inland Waters: Burlington, ON, Canada, 1986. Ongley, E. D.; Bynoe, M. C.; Percival, J. B. Can. J.Earth Sci. 1981,18, 1365-1379. Horowitz, A. J. In Chemical and Biological Characterization of Sludges, Sediments, Dredge Spoils and Drilling Muds; ASTM STP 976; Lichtenberg, J. J., Winter, J. A., Weber, C. I., Fradkin, L., Eds.; American Society for Testing and Materials: Philadelphia, PA, 1988; pp 102-113. Qualitative Analysis; Nuclepore Corp.: Pleasanton, CA, 1989. Millipore Catalog and Purchasing Guide; Millipore Products Division, Millipore Corp.: Bedford, MA, 1983. Bale, A. J.; Morris, A. W. Estuarine, Coastal Shelf Sci. 1987, 24, 253-263. Weiner, B. B. In Modern Methods of Particle Size Analysis; Barth, H. G., Ed.; John Wiley and Sons: New York, 1984; pp 135-172. Krishnappan, B. G.; Droppo, I. G.; Rao, S. S.; Ongley, E. D. Evaluation of a Filter-Fractionation Technique for Fine Sediments; NWRI Contribution 90-11; Canada Centre for Inland Waters, Burlington, ON, Canada, 1990.
Received for review February 25, 1992. Revised manuscript received M a y I , 1992. Accepted M a y 7,1992. T h e use of filter brand names does not imply a n endorsement of materials by Environment Canada.
Ozonation Byproducts: Identification of Bromohydrins from the Ozonation of Natural Waters with Enhanced Bromide Levels Joseph E. Cavanagh, Howard S. Weinberg," Avram Gold, R. Sangalah, Dean Marbury, and Wllllam H. Glare Department of Environmental Sciences and Engineering, University of North Carolina, Chapel Hill, North Carolina 27599
Timothy W. Collette, Susan D. Richardson, and Alfred D. Thruston, Jr. Environmental Research Laboratory, US. Environmental Protection Agency Athens, Georgia 306 13-7799
Introduction When ozone is used in the treatment of drinking water, it reacts with both inorganic and organic compounds to form byproducts (1). If bromide is present, it may be oxidized to hypobromous acid (21, which may then react with natural organic matter (NOM) to form brominated organic compounds (2,3). The formation of bromoform has been well documented (2, 4, 5 ) , and more recently, other byproducts, such as bromoacetic acids, bromopicrin, cyanogen bromide, bromoacetones, and bromate, have 1658 Environ. Sci. Technol., Vol. 26, No. 8, 1992
been identified (6-9). The purpose of this communication is to report the identification of bromohydrins, a new group of labile brominated organic byproducts from the ozonation of a natural water in the presence of enhanced levels of bromide. Experimental Section A sample of natural water was collected from University Lake, Orange County, NC, and used as received. The TOC of the water was 8.0 mg/L, pH 7, and the ambient bromide
0013-936X/92/0926-1658$03.0010
0 1992 American Chemical Society
