Environ. Sci. Technol. 1990,2 4 , 683-688
Spencer, W. F.; Farmer, W. J.; Cliath, M. M. Res. Rev. 1973,
National Water Well Association: Dublin, OH, 1986; pp
49, 1-40.
420-441.
Spencer, W. F.; Cliath, M. M. Soil. Sci. SOC.Am. R o c . 1970,
Marrin, D. L.; Kerfoot, H. B. Enuiron. Sci. Technol. 1988,
34, 574-578.
22, 740-745.
Peterson, M. S.; Lion, L. W.; Shoemaker, C. A. Environ. Sci. Technol. 1988,22, 571-578. Chiou, C. T.; Porter, P. E.; Schmedding, D. W. Environ. Sci. Technol. 1983, 17, 227-231. Sargent, B. P.; Green, J. W.; Harte, P. T.; Vowinkel, E. F. Open-File Rep.-U.S. Geol. Surv. 1986, No. 86-58. Steinberg, S. M.; Pignatello,J. J.; Sawhney,B. L. Enuiron. Sci. Technol. 1987,21, 1201-1208. Pignatello, J. J.; Frink, C. R.; Marin, P. A.; Droste, E. X. J . Contam. Hydrol., in press. Pignatello, J. J. Environ. Toxicol. Chem., in press. Pignatello, J. J. Environ. Toxicol. Chem., in press.
Kammer, J. A,; Smith, J. A. Water-Resour. Invest. Rep.
(US.Geol. Surv.) 88-4220
1988, No. 88-4220, 617-624.
Wershaw, R. L.; Fishman, M. J.; Grabbe, R. R.; Lowe, L. E. US.Geological Survey Techniques of Water-Resources Investigations; U.S. Geological Survey: Washington, DC, 1982; Book 5, Chapter A3. Sawhney,B. L.; Pignatello,J. J.; Steinberg,S. M. J. Environ. Qual. 1988,17, 149-152.
Weast, R. C., Ed. Handbook of Chemistry and Physics; CRC Press, Inc.: Boca Raton, FL, 1988-1989. Dilling, W. L. Environ. Sci. Technol. 1977, 2 1 , 405-409. Adamson, A. W. Physical Chemistry of Surfaces; John Wilev and Sons: New York, 1976. Brueil, C. J.; Hoag, G. E. Proceedings of Petroleum Hydrocarbons and Organic Chemicals in Ground Water: Prevention, Detection and Restoration; Houston, TX;
Received for review December 5, 1988. Revised manuscript received November 27, 1989. Accepted December 20, 1989.
Study of Copper( I I ) Association with Dissolved Organic Matter in Surface Waters of Three Mexican Coastal Lagoons Anne M. Hansen,*st James 0. Leckie,**$Enrique F. Mandelli,* and R. Scott Altmannt
Environmental Engineering and Science, Department of Civil Engineering, Stanford University, Stanford, California 94305 Copper binding by dissolved organic matter has been evaluated by potentiometric titrations on surface waters from three Mexican coastal lagoons. A copper-selective electrode technique was utilized to measure cupric ion activity. A model that accounts for the variation in binding intensity as a function of the degree of surface loading was employed to calculate the binding constants of the complex formation between cupric ion and the organic ligands in solution. Small amounts of strongly complexing ligands were found to be present in the dissolved organic fraction. These naturally occurring ligands may be responsible for the availability of some micronutrients and for the inactivation of toxic heavy metals in the studied lagoons.
Introduction Increased understanding of the composition and the concentration of dissolved organic ligands in natural waters and ligand interaction with metal ions has been the focus of numerous studies (1-5). Organic metal complexing substances have two principal sources: humic substances derived from the chemical and microbial decomposition of organic material (6) and extracellular metabolic compounds excreted from algae (7,8). The major functional groups of humic substances include carboxylic acids, phenolic and alcoholic hydroxyl groups, and keto functional groups. The structure of aquatic humic substances is unknown (9). Fulvic acids contain considerably more groups of acidic nature, particularly carboxyl and phenolic OHs, than do humic acids (10-12). Since carboxyl and hydroxyl groups of humic 'Institub de Ciencias Nucleares, Universidad Nacional Auunoma de MBxico, 04510 Mexico City, Mexico. Environmental Engineering and Science, Department of Civil Engineering,Stanford University, Stanford, CA 94305. 8 Instituto de Ciencias del Mar y Limnologia, Universidad Nacional Autdnoma de MBxico, 04510 Mexico City, Mexico.
*
0013-936X190/0924-0683$02.5010
substances are most reactive with cations, fulvic acids have a higher affinity for interactions with metal ions (13,14). It is generally agreed that fulvic acids have two general types of functional groups: salicyclic and dicarboxylic. These general types of sites may be presented in a great variety of combinations, and it is likely that no two carboxyl groups are chemically identical. Therefore, metal complexation by humic and fulvic materials must be associated by not one but a continuum of binding energies (2, 4, 15, 16).
The evaluation of interactions between metals and organic ligands present in natural waters has lately received considerable attention. Evidence for strong organic complexation has been demonstrated under controlled experimental conditions with samples from natural environments such as lakes and rivers, and with fulvic and humic acids extracted from soils (17-22). Observations of metal-binding characteristics of humic-type compounds (2),hydrous oxides (23,241,and natural sediments (25)have revealed similarities in behavior. The major similarities are the following: (1)that a variation in binding intensity is exhibited as a function of degree of site occupancy; (2) that the binding intensity for the same solid depends on the cation. This means that strong binding sites for one metal are not necessarily preferred binding sites for other metal ions; and (3) different binding energy sequences are expected for the same cations with respect to different ligand systems. Bresnahan and coworkers (20) observed a different complexing behavior of copper(I1) with soil fulvic acid than with water fulvic acid. Mantoura and colleagues (22) studied dissolved ligands from different sources and found a wide range of values for the complexing energy of any particular metal. In any case, comparison of published results should be done with much care, since the energies of binding vary as a function of the degree of loading of the ligand system. The number of protons released when a metal ion adsorbs varies with the degree of occupation of the ligand
0 1990 American Chemical Society
Environ. Sci. Technol., Vol. 24, No. 5, 1990
888
the metal-ligand association (2). The mass-balance equation for the substrate is [ S J = [SHA + WMeI
(2)
By substitution of eq l a into eq 2 , the following equation can be obtained:
For the ith ligand, the complex-formation reaction is SiHxg+ Me = SiMe + xiH
(5)
Figure 1. Map of the three lagoons and sample localization.
system. As pointed by Altmann and Leckie (161,a site may be comprised of one or more coordinating atoms, which for humic compounds are mostly acidic. Thus, as a given metal occupies the complexing sites, the proton-exchange stoichiometry is not necessarily constant. The present study was undertaken to evaluate the capacity of the dissolved organic matter to bind heavy metals in three Mexican coastal lagoons. The Mitla, Coyuca, and Tres Palos Lagoons are located on the west coast of Mexico near Acapulco (see Figure 1). These lagoons were previously described by Lankford (26). The lagoon fisheries have been impacted by the development of local agriculture and industry, and heavy-metal contamination of the lagoons could have a serious environmental impact on the local aquatic ecology. Although the present work studied interactions of dissolved organic matter and Cu(I1) by an ion-selective technique, the results are indicative of complexation to be expected by other heavy metals. The geological and physical features of the studied lagoons have been described in detail by Lankford (26). Observations indicated that dissolved organic materials found in these lagoons might have different origin. In the Mitla Lagoon (see Figure l),sampling station 1 seems to be mostly influenced by humic material from terrestrial origin, while station 3 may have a higher content originating from algae excudates. These observations were based on frequent visits to the lagoons and on observations of algal blooms and dissolved oxygen determinations. The Mitla Lagoon has been isolated from the sea since 1964 (27) and consequently has very poor circulation, is highly eutrophied, and has an accelerated sedimentation rate. The Coyuca and Tres Palos Lagoons are intermittently open to the sea. However, like Mitla, they are hyposaline year-round.
Outline of the Model Assuming a ligand to metal stoichiometry of 1:1, the overall formation reaction can be represented as follows: SH, + Me = SMe + ZH (1)
K=
[SMe][H]" [SH,I [Me1
(la)
where SH, represents the mixture of chelating sites, Me the metal ion, SMe the metal-ligand association, and ZH the average number of protons released during the process of chelation. The equilibrium value, K, is obtained for the mixture of sites available for chelation. Therefore, R is not a constant, but a function that varies depending upon the solution composition. Thus, K cannot be used to uniquely describe the stoichiometry and the intensity of 684
Environ. Sci. Technol., Vol. 24, No. 5, 1990
(4)
where Ki is the equilibrium constant for the ith ligand of the natural mixture of polymer fulvic acid. Ki is a very useful value that can be used to quantify the energies and stoichiometries associated with the complex formation (2). However, Ki cannot be measured directly. It is only possible to measure average values of the complex formation reaction such as, for example, average proton-exchange stoichiometry. There are basically two approaches to estimate this quantity: One is the batch addition of metal ion and titration of the solution in the pH range of interest (not pH stated). The other method is incremental titration with the metal ion, controlling the hydrogen ion activity with a pH stat and recording incremental additions of acid and base. The Scatchard total ligand-metal site approach assumes one or a few different types of ligand sites and constant K values (17). However, in the case of complex mixtures of naturally occurring organic ligands, simple models utilizing smaller molecules of known and well-defined substances cannot be applied (2,20,28). Perdue and Lytle ( 4 ) suggested that constants obtained from the Scatchard equation must be regarded as empirical curve-fitting parameters with no chemical significance and that these lumped parameters represent classes of related ligands rather than discrete ligands. An extended Scatchard approach has included interactions between sites, i.e., the binding to one site alters the affinity of the macromolecule for successive binding. Since, in the absence of adequate information it is not possible to derive a mathematical expression for the dependence among sites, Cantor and Schimmel (29) introduced an average-based constant that describes these interactions. Other modeling approaches take into account the variation in metal-ligand binding intensity as a function of system composition. At low site occupancy and within a narrow range of system variables, a fixed stoichiometry can be used to describe the metal-ligand system, and constant binding energetics can be assumed (30, 31). Gamble and co-workers (2) initiated a mathematical development of their metal-ligand model from eq 1 and 2 , but did not include varying proton-exchange stoichiometry. The mass balance equations for the bound ligand and the free ligand, respectively, are n
[SMe] = C[SiMe] i=O
n
[SH,I =
C [S&I
i=O
(7)
Equations 6 and 7 can be substituted into eq la, giving
Table I. Chemical Composition of Samples
lagoon Mitla
c [S,HZiI
Equation 8 shows that K is directly related to the individual equilibrium constants of the n components, being the weighted average of all these components. The mass balance for the complexing sites is given by eq 2 and the mole fraction of the ith component is
5 9.6 9.7 0.7
1810 180 24.9 108 18.9
1850 246 34.5 118 27.7
+
-91
Assuming a continuous distribution of sites and replacing the summation by an integral yields
-11
~
-7
,
,
,
-6.6
, -6.2
and by differentiation
The characteristics of K are the following: K is a differential equilibrium function that is useful for the quantification of the energy involved and the distribution of sites during the formation of chelates within complex ligand system. K can be used for calculation of thermodynamic functions and compared to single ligand systems. By definition the sum of the mole fractions of the free and bound ligand is unity: (13)
and by differentiation (14)
Substituting eq 14 into eq 1 2 gives
The I? values can be estimated from experimental results, and K can be estimated as the slope of the plot K versus %Me*
Experimental M e t h o d o l o g y
Five water samples were collected from the surface of three lagoons (Figure 1) and filtered through 0.45-pm Millipore filters. The samples were stored in acid-leached Pyrex containers in the dark under refrigeration until analyzed. Chemical analyses for the following constituents
,
, -5.8
,
,
,
,
,
,
-5
-5.4
-4.6
,
, -4.2
Log [ C " ] t O t 0
UV-lrradlated
Flgure 2. Cu2+ titration of
dXSHi. = -dXSMe
4 9.4 5.0 0.2
-4
Introducing eq 7 and 9 into eq 8 yields the following:
=
Tres Palos
1 2 3 sample pH in situ 9.8 10.6 8.7 DOC, mg/L 22.2 19.3 20.5 DOC in irrad 1.0 0.8 0.4 samples, mg/L C1-, mg/L 1560 1740 1710 SO-:, mg/L 183 189 186 32.8 30.7 29.8 Ca2+,mg/L 111 107 107 Mg2+,mg/L 45.3 72.0 71.4 suspended matter, me/L
i=O
XSH,- -k XSMe
Coyuca
+
natura1
untreated and UV-irradiated sample 1.
were conducted on filtered samples: C1-, SO-,: Ca2+,and Mg2+. Portions of the filtered samples were exposed to high-intensity UV irradiation by means of a 1200-W mercury arc tube photooxidation unit (Hanovia Engelhardt 189 A), and dissolved organic carbon (DOC) was determined in both filtered and irradiated samples on a Dohrman carbon analyzer. The results of these analysis are presented in Table I. A copper-selective electrode titrimetric method was used (3). This technique has been restricted to organic matter-Cu(I1) interactions in fresh waters because of the interference of the chloride ion. In the present work, the concentration of chloride ion varied from 0.044 to 0.052 M in the five samples (see Table I). In this interval, the highest permitted Cu(I1) concentration was 5.92 X lo4 M (32). The total copper concentrations in these experiments were all below this critical value. Cupric ion activity was measured with an Orion Model 94-29 cupric electrode against an Orion Model 90-01 single-junction reference electrode using a digital pH/mV meter Model Orion 701 A. The pH was measured with an Ingold Glass electrode Model HA 201 against a similar reference electrode. Cupric ion was added to the water samples by means of a microburet in 30 steps in aliquots between 0.01 and 0.25 mL to cover the interval between lo-' and M total copper. Titrations were performed at pH 6 and 7. A solution of double-distilled nitric acid was used to bring the samples to the initial pH value. All titrations were carried out under N2(gas)atmosphere in a thermally controlled borosilicate beaker at 25 "C. A simple acid-base titration with carbonate-free KOH was performed on each sample. The pH was brought to 3 by addition of a solution of twice-distilled HNO,, and the resulting solution was titrated with carbonate-free KOH solution to pH 10.5. The equivalence point was Environ. Sci. Technol., Vol.
24, No. 5, 1990 665
Table 11. Calculation of Protons Released per Copper Bound
sample protons, equiv/L L t l , equiv/L protons/DOC, equiv/g [L,I/[DOCl, equiv/g protons/ L,,
2 2.9 x 3.0 x 1.5 x 1.6 x 0.9
1
2.1 x 2.9 x 0.9 x 1.3 x 0.7
10-5 10-5 10-3 10-3
3 2.4 x 3.1 x 1.2 x 1.5 x 0.8
10-5 10-5 10-3 10-3
5 2.5 x 2.6 x 2.6 x 2.7 x 1.0
4
2.7 x 2.3 x 5.4 x 4.6 x
10-5 10-5 10-3 10-3
10-5 10-5 10-3
10-3
1.2
10-5 10-5 10-3 10-3
Table 111. Selected Values of K and Ki at pH 6 and 7
XCUL
pH 6.0 0.04 0.06 0.16 0.29 0.44 0.60 pH 7.0 0.04 0.06 0.16 0.30 0.43
K 14.1 7.9 2.4 0.9 0.5 0.4 0.6 0.3 0.1 0.05
0.04
samde 3
2
1
K
Ki
K
Ki
884 132
9.6 5.8 1.6
327 79.5 19.4 1.9
4.0 2.7
0.3
0.5 0.3 0.2
68.5 29.0 9.0 1.5 0.6 0.3
8.9 7.2 1.2 0.27 0.12
0.7 0.5 0.2 0.06 0.04
15.4 5.7 1.6 0.19 0.07
21.3
4.3 1.2 0.5
0.6
0.4 0.3
18.7
0.7
4.4 0.5
0.5
0.17 0.08
0.1 0.07 0.04
1.1
calculated by the Gran method (33),and the difference between the blank and the natural samples gave an estimate of the number of protons released per milligram of dissolved organic carbon.
Results and Discussion In the determination of empirical association constants between metallic ions and dissolved or suspended ligand systems, it is important to have a reference of similar chemical composition. This was accomplished by UV treatment of the samples, and Figure 2 shows titration curves of untreated and UV-irradiated sample 1, where pCu2+ is plotted against total Cu. The major part of complexation of the metal ion can be attributed to the presence of organic material (Figure 2). Data from the UV-irradiated sample are practically parallel to the data obtained by titration of a solution of KN03 at the same ionic strength (not shown in the figure). The cupric ion is nearly 100% associated with the organic material until a total copper concentration of 5 X lo-' M. The percent of bound copper decreases with increasing concentration of total copper and finally approaches data for the UVirradiated sample. There are no generally agreed upon methodologies for estimating total ligand concentrations for complex materials as humic substances. All methods for this estimation suffer some criticism (2). We selected a rather simple technique for estimating this value. The result allowed us to feed the model that describes varying binding constant with varying metal ion concentration. Figure 3 illustrates the method used by Van den Berg and Kramer (17) to estimate total concentration of Cu(11)-associatingsites in the system. According to eq 3, by plotting [Cu2+]/[CuL](L= ligand) versus [Cu2+],the slope gives the inverted value of the total ligand concentration. The difference in equivalents of the titrations of each sample is shown in Table 11. Protons released per copper bound is close to 1 for all of the samples, and 1:l complexation between the cupric ion and the organic ligand was assumed in each case (see Table 11). Complex formation between Cu(I1) and the inorganic ions in the samples was negligible at pH 6. However, at 686
4
Ki
Environ. Sci. Technol., Vol. 24, No. 5, 1990
1.1
5
K
Ki
K
Ki
1.8
7.8 4.6 0.9 0.9 0.6
1.4 0.6 0.2 0.2 0.2
10.0 3.7
0.04 0.03
0.04 0.03 0.02 0.01
1.1
0.8 0.7 0.5
0.01 0.01 0.008 0.006
0.03
0.007
1.1
0.2 0.2 0.51 0.20 0.09
0.04
perimental pH values. This result was expected since cupric ions compete with other metals, including protons, for the binding sites of the organic material. When not occupied by cupric ion, these functional groups must be occupied by protons or by other cations. Nonconstant binding energies were found as fractional binding with metal increased a t a constant pH. This behavior is attributable to the common theory that polyligand materials are composed of various types of sites with a range of binding energies (4, 16). Humic substances exhibit polyfunctional character and heterogeneous acidic functional group chemistry. Therefore, these compounds present a broad spectrum of reactivity toward metal ions (36). Conclusions
The natural water samples from the three Mexican coastal lagoons were very rich in organic material, and no prior treatment nor preconcentration had to be accomplished that could affect the chemical integrity of these substances. The sources of organic material in the studied aquatic systems include autochthonous primary production and allochthonous material entering through draining. The contribution from each source may vary widely in time and space. However, exudates of cells are likely to be important ligands for trace metals in waters where dense phytoplankton blooms occur, as in the case of the eutrophied Mitla Lagoon. Although DOC and L,, values were rather similar in the three sampling stations in the Mitla Lagoon, the Kishowed much higher values for sample 1a t low cupric ion loading. This is an indication of the presence of small amounts of strongly complexing sites in this area, which are mostly introduced by the humic acid input from terrestrial origin. The calculated complex-formation constants show that organic chelates play a significant role in the speciation of Cu(I1) in the studied aquatic systems. These organic compounds exhibit polyfunctional character and heterogeneous acidic functional group chemistry and therefore present a broad spectrum of reactivity toward the cupric ion. Metal complexation in such complex ligand systems is often assumed to have a single type of bonding dominate. However, natural polyligands like humics, hydrous oxides, clays, and bacterial surfaces have no constant reaction stoichiometry nor constant reaction energetics. The theoretical analysis of these systems has been delayed by its complexity. Constants obtained from the Scatchard equation must be regarded as empirical curve-fitting parameters with no chemical significance. These constants represent classes of related ligands rather than discrete ligands. A limited number of measured values pertaining to a narrow range of solution composition cannot be used to draw general conclusions about the complex formation between metal ions and the dissolved organics for even a single humic substance. Also, the metal-organic complex formation constants must be used with caution, because they vary as a function of the solution composition. It is observed that metal binding intensity is a function of the system composition, so the sites being occupied a t high adsorption densities are different from those being occupied at low adsorption density, forming intrinsically weaker adsorption bonds. The present modeling approach takes into acocunt the variation in metal-ligand binding intensity as a function of system composition. This method to evaluate Cu(11)-organic matter interactions is the most appropriate
as long as the chemical nature of the dissolved organic substances is not perfectly well-known. Registry No. Cu, 7440-50-8;C, 7440-44-0; Ca, 7440-70-2;Mg, 7439-95-4.
Literature Cited Buffle, J.; Deladoey, P.; Greter, F. L.; Haerdi, W. Anal. Chim. Acta 1980,116, 255-274. Gamble, D. S.; Underdown, A. W.; Langford, C. H. Anal. Chem. 1980,52, 1901-1908. McKnight, D. M. M.Sc. Thesis, Massachusetts Institute of Technology, Cambridge, MA, 1978. Perdue, E. M.; Lytle, C. R. Environ. Sci. Technol. 1983, 17,654-660. Li, Y.-H.; Burkhardt, L.; Tereoka, H. Geochim. Cosmochim. Acta 1984, 48, 1879-1884. Jackson, K. S.; Jonasson, I. R.; Skibben, G. B. Earth-Sci. Rev. 1978, 14,97-146. Swallow, K. C.; Westall, J. C.; McKnight, D. M.; Morel, N. M. L.; Morel, R. M. M. Limnol. Oceanogr. 1978,23,538-542. McKnight, D. M.; Morel, F. M. M. Limnol. Oceanogr. 1979, 24, 823-837. Thurman, E. M. Organic Chemistry of Natural Waters; Martinus Nijhoff/Dr. W. Junk Publishers: Dordrecht, The Netherlands, 1985; Chapter 10, pp 273-361. Stevenson, F. J.; Butler, J. H. A. In Organic Geochemistry; Eglinton, G., Murphy, M. T. J., Eds.; Springer Verlag: Berlin, 1969; pp 534-557. Schnitzer, M.; Skinner, S. I. M. Soil Sci. 1965,99,278-284. Gamble, D. S.; Schnitzer, M. In Trace Metals and Metal-Organic Interactions in Natural Waters;Singer, P. C., Ed.; Ann Arbor Science Publishers: Ann Arbor, MI, 1973; pp 265-302. Gamble, D. S. Can. J . Chem. 1970,48, 2662-2669. Gamble, D. S. Can. J . Chem. 1972,50, 2680-2690. Sposito, G. In Chemistry in the Soil Environment; American Society of Agronomy and Soil Science: Madison, WI, 1982. Altmann, R. S.; Leckie, J. 0. In Oceanic Processes in Marine Pollution; O’Connor, T. P., Burt, W. V., Duedall, I. W., Eds.; Robert E. Krieger Publishing Co.: Malabar, FL, 1987; Vol. 2, Chapter 13, pp 145-157. Van den Berg, C. M. G.; Kramer, J. R. Anal. Chim. Acta 1979, 106, 113-120. Van den Berg, C. M. G.; Kramer, J. R. In Copper in the Environment, Part 1, Ecological Cycling; Nriagu, J. O., Ed.; Wiley: New York, 1980; Chapter 6. Stevenson, F. J. Soil Sci. 1977, 123, 10-17. Bresnahan, W. T.; Grant, C. L.; Weber, J. H. Anal. Chem. 1978, 59, 1675-1679. Buffle, J.; Greter, F. L.; Haerdi, W. Anal. Chem. 1977,49, 216-222. Mantoura, R. F. C.; Dickson, A.; Riley, J. P. Estuarine Coastal Mar. Sci. 1978, 6, 387-408. Benjamin, M. M.; Leckie, J. 0. Contaminants and Sediments; Baker, R. A., Ed.; Ann Arbor Science Publishers: Ann Arbor, MI, 1980; Vol. 2. Benjamin, M. M.; Leckie, J. 0. J. Colloid Interface Sci. 1981,83, 410-419. Lion, L. W.; Altmann, R. S.; Leckie, J. 0. Enoiron. Sci. Technol. 1982, 16, 660-666. Lankford, R. R. In Estuarine Processes; Wiley, R., Ed.; Academic Press: New York, 1977; Vol. 2, p 182. Paez-Osuna, F. Instituto de Ciencias del Mar y Limnologia, Universidad Nacional Autdnoma de Mgxico, personal communication, 1981. Gamble, D. S.; Schnitzer, M.; Hoffman, I. Can. J. Chem. 1970, 48, 3197-3204. (29) Cantor, C. R.; Schimmel, P. R. Biophysical Chemistry; W. H. Freeman Co.: San Francisco, CA, 1980; Vol. 111, pp 849-886. (30) James, R. 0.;Healy, T. W. J . Colloid Interface Sci. 1972, 63, 480-499. (31) Davis, J. A.; James, R. 0.;Leckie, J. 0. J. Colloid Znterface S C ~1978, . 63, 480-499. Environ. Sci. Technol., Vol. 24, No. 5, 1990 687
Environ. Sci. Technol. 1990, 24, 688-691
Orion Research, Inc., Instruction Manual, Cupric Electrode Model 94-29, 1979. Gran, G. Acta Chem. Scand. 1950, 550-577. SillBn, L. G.; Martell, A. E. Stability Constants of Metal Ion Complexes. Spec. Publ.-Chem. SOC.1971, No. 25. Buffle, J. Anal. Chim. Acta 1980, 118, 29-44.
(36) Sposito,G. Presentation at the IHSS meeting,Birmingham, UK, July, 1984. Received for review September 4, 1987. Revised manuscript received March 29, 1988. Accepted June 22, 1989.
Aspiration Efficiency: Unified Model for All Forward Sampling Angles Sun11 Hangal and Klaus Willeke' Aerosol Research Laboratory, Department of Environmental Health, University of Cincinnati, Cincinnati, Ohio 45267-0056
We have developed a unified aspiration efficiency model for sampling with round inlets at 0-90' forward angles from horizontal aerosol flows. The aspiration efficiency is represented by a single equation as a function of the sampling angle, the ratio of wind to inlet velocity, and different inertial parameters for 0-60' and 45-90' sampling. The equation for 45-90' sampling is an extension of the Laktionov equation for 90'. The new model fits the experimental data within experimental accuracy.
Introduction When airborne particles are sampled in ambient and industrial environments through an inlet, three processes may change the original aerosol concentration and size distribution: (1) aspiration to the face of the inlet, (2) bounce from the front edge of the inlet, (3) transmission losses in the inlet. The changes that occur during sampling must be assessed quantitatively so that sampling errors can be compensated for. The overall sampling efficiency, E,, can thus be represented by the product of three distinct efficiencies:
E , = EaErEt
(1)
where E , is the aspiration efficiency, which is the ratio of particle concentration at the inlet face to the particle concentration in the undisturbed environment. E, is the entry efficiency, which is the ratio of particle concentration passing the inlet face to the particle concentration incident to the inlet face. Et is the transmission efficiency, which is the ratio of the particle concentration exiting the inlet to the particle concentration just past the inlet. Whenever possible, inlets are designed with sharp edges to that particle bounce is negligible. The sharp-edge inlets are designed so as to meet the design criteria of Belyaev and Levin (I),which specify that for a sharp-edged inlet (a) the ratio of outer to inner inlet diameter is 61.1 and (b) if the ratio of outer to inner inlet diameter is 31.1,the ratio of inlet thickness to inlet inner diameter is 60.5 and the angle of taper is 615'. Thus, E , = E,E, (2) The overall sampling efficiency and transmission efficiency have been measured in our wind tunnel for a wide range of wind and inlet velocities, sampling angles, and round sharp-edged inlet sizes (2-7). The aspiration efficiency values, calculated from these data through eq 2, have been used for the development of our unified model for aspiration efficiency. The wind condition in the ambient and in wind tunnel systems is turbulent. The effect of turbulence intensity and scale on aspiration efficiency of sharp-edge inlets was found to be negligible at isokinetic (8) and anisokinetic conditions (9). In our studies on the effect of turbulence 688
Environ. Sci. Technol., Vol. 24, No. 5, 1990
(7), we found that the turbulence intensity and scale influenced particle deposition inside the inlet because turbulent motion is superimposed on the convective flow. Our data indicated that turbulence affected the transmission efficiency for small inlets but had negligible effect on the overall sampling efficiency, from which we also conclude that the effect of turbulence on aspiration efficiency is negligible. In our wind tunnel, the turbulence intensity varied from 1.4 to 7.5% and the scale ranged from 0.5 to 10 cm, which are close to the range of values used in the other studies on the effect of turbulence (8,9).The effect of turbulence has been ignored in the studies on aspiration (1, 10-16). Aspiration for round sharpedged inlets has been studied extensively (1, 10-16) and several equations have been developed to describe it. Comparison of these equations with our wind tunnel data has shown that some of these equations agree with our data over specific ranges of sampling angle, but none of these equations agree over the entire range of 0-90' (4-6), where the sampling angle is the angle at which the inlet is oriented with respect to the horizontal wind direction (2-7).
Available Aspiration Equations Studies on aspiration ( I , 10-16) have shown that the aspiration efficiency depends on the inertial behavior of the particles, the flow conditions, and the sampling angle, 0. Therefore, E, = R, e) (3) where Stokes number, St, and velocity ratio, R, are St = TV,/D~ (4)
m,
R = U,/Ui Particle relaxation time, T , is
(5)
(6) where p is the particle density, p is the air viscosity, Di is the infet diameter, Uwis the wind velocity, and Viis the inlet velocity. Belyaev and Levin (1)developed the following equation for isoaxial sampling from experimental data obtained by flash illumination photography: E, = 1 + ( R - 1)[1- 1/[1 + (2 + 0.617/R)St]] (7) valid for 0.16 C R C 5.6, 0.18 < St < 2.03. Davies and Subari (10) developed aspiration efficiency equations for sampling at 0 and 90' based on experimental data at R C 1. The Davies and Subari equation at 0' is E, = St(R)3/2(4R+ 0.62) 1 - (1 - R) O.lSt(l - R) + (R)3/2[l+ St(4R + 0.6211 T
0013-936X/90/0924-0688$02.50/0
= ppdp2/ 18p
(8) 0 1990 American Chemical Society
