Environ. Sci. Technol. 2002, 36, 2806-2814
Comparison of the Molecular Mass and Optical Properties of Colored Dissolved Organic Material in Two Rivers and Coastal Waters by Flow Field-Flow Fractionation ELIETE ZANARDI-LAMARDO,* CATHERINE D. CLARK,† CYNTHIA A. MOORE, AND ROD G. ZIKA Division of Marine and Atmospheric Chemistry, Rosenstiel School of Marine and Atmospheric Science, University of Miami, 4600 Rickenbacker Causeway, Miami, Florida 33149
Colored dissolved organic material (CDOM) is an important sunlight absorbing substance affecting the optical properties of natural waters. However, little is known about its structural and optical properties mainly due to its complex matrix and the limitation of the techniques available. A comparison of two southwestern Florida rivers [the Caloosahatchee River (CR) and the Shark River (SR)] was done in terms of molecular mass (MM) and diffusion coefficients (D). The novel technique Frit inlet/frit outlet-flow field-flow fractionation (FIFO-FlFFF) with absorbance and fluorescence detectors was used to determine these properties. The SR receives organic material from the Everglades. By contrast, the CR arises from Lake Okeechobee in central Florida, receiving anthropogenic inputs, farming runoff, and natural organics. Both rivers discharge to the Gulf of Mexico. Fluorescence identified, for both rivers, two different MM distributions in low salinity water samples: the first was centered at ∼1.7 kDa (CR) and ∼2 kDa (SR); the second centered at ∼13 kDa for both rivers, which disappeared gradually in the river plumes to below detection limit in coastal waters. Absorbance detected only one MM distribution centered at ∼2 kDa (CR) and 2.2-2.4 kDa (SR). Fluorescence in general peaked at a lower MM than absorbance, suggesting a different size distribution for fluorophores vs chromophores. A photochemical study showed that, after sunlight, irradiated freshwater samples have similar characteristics to more marine waters, including a shift in MM distribution of chromophores. The differences observed between the rivers in the optical characteristics, MM distributions, and D values suggest that the CDOM sources, physical, and photochemical degradation processes are different for these two rivers.
Introduction An important and abundant form of organic matter in the ocean is dissolved organic material (DOM), which is involved * Corresponding author e-mail:
[email protected]; phone: (305)361-4713; fax: (305)361-4894. † Present address: Department of Environmental and Chemical Sciences, Chapman University, One University Drive, Orange, CA 92866. 2806
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 13, 2002
in chemical and biological processes in the oceanic carbon cycle (1). The main sources of DOM in coastal marine waters are river inputs and in situ production by phytoplankton, which contribute different compounds chemical to the DOM (2). This in turn may affect its reactivity, structural, and optical characteristics (2). Colored dissolved organic material (CDOM) refers specifically to the fraction of DOM that absorbs visible and near-ultraviolet radiation. CDOM imparts color to the ocean, and it is involved in photochemistry in surface waters. When excited by sunlight, CDOM fluoresces (1, 3) and undergoes a complex series of reactions. Photodegradation experiments showed that some components of DOM are more resistant to photolysis than others, since the loss of DOM was slower than the observed loss of absorptivity of CDOM (4-6). Photodegradation of CDOM also produces labile low molecular mass (MM) carbon compounds, which stimulate the growth of phytoplankton and bacteria (2, 7). Some studies have demonstrated that samples irradiated prior to bacterial degradation resulted in an increased decomposition when compared to biological activity alone (4, 7). However, a mass balance calculation demonstrated that higher MM compounds also become biologically available during sunlight exposure (7). From these studies, it seems that photochemical processes producing DIC rather than biological activity producing CO would be the mechanism that controls DOM residence time in waters. Microbial activity together with photochemistry are considered important mechanisms for removal of organic carbon from the ocean (4, 8). Since photochemical and biological transformation processes are expected to result in lower molecular mass material, a decrease in the MM from the freshly input CDOM in rivers to more aged material in coastal waters is also expected. Measuring the MM of large macromolecules such as DOM is challenging, mainly because it has a very complex matrix. A number of techniques including size exclusion chromatography, ultrafiltration, light scattering, ultracentrifugation, vapor pressure osmometry, gel permeation chromatography, and mass spectrometry have been used to measure the MM of humic and fulvic acids, which form a significant fraction of CDOM. Reported values for macromolecules determined by these methodologies range from 0.50 to greater than 10 kDa (6, 9-16). One possible reason for this wide range of values is artifacts of the techniques used, such as ion exclusion, ion exchange, adsorption, and aggregation (12). A more recent approach to separating humic materials by molecular size is flow field-flow fractionation (FlFFF). This refers to a chromatographic analytical separation technique carried out in an unpacked ribbon-shaped channel with a semipermeable membrane layered at one side of the channel (17). In an FlFFF channel, separation depends only on the diffusion coefficient (D) of the sample, i.e., compounds will be separated by molecular size. FlFFF has several advantages as compared to other DOM fractionation techniques such as ultrafiltration, gel permeation, and size exclusion chromatography, which may suffer from adsorption and charge repulsive effects (10, 13). It is a nondestructive method that allows for the use of multiple in-line detectors and for sized fractions to be collected for analysis by other techniques. FlFFF has previously been used in studies of water-soluble polymers (18), protein characterization (19), and dextran carbohydrates (20). Colored humic substances from soils, rivers, lakes, and groundwater have been analyzed by coupling FlFFF to UV absorbance detectors, giving MM distributions ranging from 3 kDa (13), and the Baltic Sea (23). Since CDOM is a very complex mixture of organic materials, many chromophores and fluorophores may occur within the same MM distribution. For both rivers, the chromophores’ size distribution was centered at a larger MM than the fluorophores. This suggests that some of the absorbing moieties either do not fluoresce or fluoresce less intensely than the lower MM compounds. The same result was also verified by recent studies of natural organic material (40) and humic substances (9) where authors observed smaller molecules fluorescing more intensively than the larger ones. Specht et al. (40) concluded that larger molecules could go from the excited state to the ground state with no light emission. It is well-known that measurements of absorbance may include some light scattering contribution, which in turn would shift the chromophores peak maxima to larger molecular masses. However, it is believed that, if any, this effect is minimal in FlFFF analysis because the samples are filtered right before the injection and the amount of the fractionated CDOM on the detector is very small. Perhaps most significantly, the CR had a lower MM distribution than SR for all stations for the first size fraction with chromophores centered at 2.02 ( 0.04 kDa (CR) and ranging from 2.40 to 2812
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 13, 2002
2.12 ( 0.06 kDa (SR). Fluorophores were centered at 1.73 ( 0.03 and 1.98 ( 0.06 kDa for CR and SR, respectively. The size dependency of fluorescence properties might also be the reason for the higher fluorescence yield of CR as compared to SR (9, 40). Another interesting difference between the rivers was that samples from the SR shifted to a lower MM distribution of the chromophore from the river into coastal waters, but this did not occur in the CR samples. These observations together suggest that there are some structural differences between the CDOM in the rivers. These differences could be due to a combination of any of the following factors: different input sources, different rates of introduction of fresh organic material, different dilution rates of the fresh river water by coastal waters with low DOM concentration, and different exposure times to photochemical and biological transformation processes, resulting in molecules with different size distributions. The two rivers receive different inputs of organic material: the CR arises from Lake Okeechobee and passes through farming and urbanized regions, potentially receiving CDOM from different sources such as agriculture, oil fuel combustion, and vegetation around the river. The SR is in a mangrove coastline in the Everglades National Park, a protected pristine area with little anthropogenic contributions and constant introduction of CDOM from the coastal vegetation. From Figure 2, the CR had a higher fluorescence than the SR for the same absorbance intensity, suggesting that the organic material is different. Since the CR is much longer (about 60 nautical miles from Lake Okeechobee), photolabile material could have already been degraded at the freshest sampling station, resulting in a lower MM distribution. This lower MM distribution may be more resistant to photodegradation (predominantly refractory material) but might be photobleaching and diluting with marine waters (no shift in MM with a decrease in optical intensities). The lower CR is deeper than SR, and increased water column mixing may decrease the sun exposure time of the compounds, delaying the degradation process. In addition, the preferential sorption of larger sized humic matter to inorganic particles (40) and subsequent sedimentation in the CR during the longer passage to the sea could explain the smaller size of CDOM. Another possibility for the differences in size distribution between the rivers is that the CDOM input from the Everglades might contain a higher proportion of larger sized humic material. Since SR receives a continuous addition of this “fresh” organic material as it passes through mangrove swamps, the MM distribution may well be expected to be higher, since photodegradation has not yet occurred. Even over a short distance in SR (sampling range about 9 nautical miles), it was possible to see a shift in the MM distribution of the chromophores. All these factors together suggest that sources, physical (sedimentation and dilution), and degradation processes (photo or biological) contribute to the MM distribution of CDOM in natural waters. To evaluate the effect of mixing, a freshwater sample (CR no. 55) was brought to the same salinity as the most saline sample (S ) 33.5) with filtered Gulf Stream water [salinity ) 36.2, absorbance (350 nm) ) 0.0001, CDOM fluorescence ) 0 QSU, and TOC ) 94 µM C]. The decrease in the fluorescence and absorbance intensities in the diluted bulk samples was similar to those found in the CR sequence. The second size fraction from the FlFFF analysis completely disappeared (based on the loss of optical properties), and the MM distribution remained the same. This suggests that dilution in the CR river plume is the significant factor in the changes observed. To investigate the effect of sunlight with no dilution in the river plume, a photochemical study was performed. A fresh river sample from the CR sequence (no. 55; salinity ) 1.3) was exposed directly to natural sunlight for 7 days in a
TABLE 5. Optical Properties and MM Distribution Based on the Peak Maxima for a Fresh CR Sample (No. 55, Salinity 1.3) before and after Sun Exposure bulk samples control sun exposed % decay
Abs (330 nm)
CDOM Fl (QSU)
0.190 0.031 84
273.2 43.6 84
first fraction
Abs (330 nm)
CDOM Fl (350/450 nm)
control sun exposed % decay
10.6 1.6 85%
3.3 0.2 94%
MM Abs (kDa)
MM Fl (kDa)
first fraction control sun exposed % decay
2.18 1.96 10
1.82 1.79 2
quartz tube. Absorbance (200-800 nm) and fluorescence (354/496 nm) were measured in the bulk water samples both before and after sun exposure. The decrease in the bulk optical properties was ∼84% (Table 5), similar to that detected in the fractions by FlFFF analyses (see below) and to data obtained in the CR natural water sample sequence, i.e., 70% decay for fluorescence and 91% for absorbance (Table 1). After irradiation, the river sample fractogram was similar to that obtained for coastal water (Figure 7a,b). Fluorescence intensity decreased 94% with no shift in MM distribution, and absorbance signals were 85% lower with a 10% shift in MM (Table 5, Figure 7a,b). The dark control showed no changes. Some recent studies have reported a shift to lower MM distribution of humic substances after irradiation (5, 6). In fact, Monsallier et al. (5) also observed a decrease on the center of MM distribution for fluorophores, which was not verified in this study. The results found in this experiment would suggest that the fraction of CDOM containing the absorbing moiety is undergoing degradation to lower MM material by photochemical processes, but this time scale or energy applied probably was not enough to breakdown the fraction of CDOM containing the fluorophores. This indicates that degradation in the SR river plume is a potential factor to explain the changes observed together with the mixing occurring in the area. These results suggest that photochemical processes are an important mechanism in altering the molecular and optical characteristics of CDOM in surface waters during the flow of the source waters toward to the sea. On the basis of the results obtained in this study, mixing is the dominant process in the CR, and both mixing and photodegradation processes are simultaneously occurring in the SR samples. Evidence for different sources and physical and chemical processes have been found between the river transects studied. Apparent molar absorptivites () suggested different sources of CDOM including higher plants, phytoplankton, and bacteria in fresh to more saline waters. Differences in MM distribution and diffusion coefficients (D) between the rivers indicate that different kinds of material are entering and/or different processes or at different rates are occurring. Interestingly, for both rivers, fluorophores were centered at lower MM distribution than the chromophores, suggesting that some absorbing moieties either do not fluoresce or fluoresce less intensively than the lower MM compounds. Even though some improvement is expected for CDOM analysis, FIFO-FlFFF demonstrated to be a powerful tool for elucidating MM distribution and CDOM characteristics in fresh and marine environments.
FIGURE 7. (a) Fluorescence (350/450 nm) as a function of molecular mass (kDa) in a photodegradation study of a fresh CR sample (no. 55, salinity 1.3). Before (s) and after (‚‚‚) 7 d of sunlight exposure. The flow rates used were channel (sample + frit) ) 1.40 mL min-1 and cross-flow ) 4.0 mL min-1. (b) Absorbance (330 nm) as a function of molecular mass (kDa) in a photodegradation study of a fresh CR sample (no. 55, salinity 1.3). Before (s) and after (‚‚‚) 7 d of sunlight exposure. The flow rates used were channel (sample + frit) ) 1.40 mL min-1 and cross-flow ) 4.0 mL min-1.
Acknowledgments The authors gratefully acknowledge support of this work by the Office of Naval Research (ONR N000149810625 and N000140110030). E.Z.L. thanks CNPq, the Brazilian Graduate Research Fellowship Program. The authors thank the crew of the R/V Calanus, Bob Chen, Charles Farmer, Erik Stabenau, and Raphael Tremblay for assistance with sample acquisition and associated measurements.
Literature Cited (1) Zepp, R. G. In Marine Chemistry, An Environmental Analytical Chemistry Approach; Gianguzza, A., Pelizzetti, E., Sammartano, S., Eds.; Kluwer Academic Publishers: Dordrech, The Netherlands, 1997; p 329. (2) McKnight, D. M.; Aiken, G. R. In Aquatic Humic Substances: Ecology and Biogeochemistry; Hessen, D. O., Tranvik, L. J., Eds.; Springer-Verlag: Berlin and Heidelberg, 1998; Chapter 1. (3) Coble, P. G. Mar. Chem. 1996, 51, 325. (4) Moran, M. A.; Sheldon, W. M., Jr.; Zepp, R. G. Limnol. Oceanogr. 2000, 45 (6), 1254. (5) Monsallier, J. M.; Scherbaum, F. J.; Buckau, G.; Kim, J.; Kumke, M. U.; Specht, C. H.; Frimmel, F. H. J. Photochem. Photobiol. A 2001, 138, 55. (6) Schmitt-Kopplin, P.; Hertkorn, N.; Schulten, H. R.; Kettrup, A. Environ. Sci. Technol. 1998, 32, 2531. VOL. 36, NO. 13, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
2813
(7) Miller, W. L.; Moran, M. A. Limnol. Oceanogr. 1997, 42 (6), 1317. (8) Miller, W. L.; Zepp, R. G. Geophys. Res. Lett. 1995, 22 (4), 417. (9) Artinger, R.; Buckau, G.; Geyer, S.; Fritz, P.; Wolf, M.; Kim, J. I. Appl. Geochem. 2000, 15, 97. (10) Beckett, R.; Jue, Z.; Giddings, J. C. Environ. Sci. Technol. 1987, 21, 289. (11) Chin, Y. P.; Aiken, G.; O’Loughlin, E. Environ. Sci. Technol. 1994, 28, 1853. (12) Chin, Y. P.; Gschwend, P. M. Geochim. Cosmochim. Acta 1991, 55, 1309. (13) Pelekani, C.; Newcombe, G.; Snoeyink, V. L.; Hepplewhite, C.; Assemi, S.; Beckett, R. Environ. Sci. Technol. 1999, 33, 2807. (14) Reid, P. M.; Wilkinson, A. E.; Tipping, E.; Jones, M. N. Geochim. Cosmochim. Acta 1990, 54, 131. (15) Thurman, E. M. Aquatic Humic Substances. Organic Geochemistry of Natural Waters; M. Nijhoff and W. Junk: Dordrecth, The Netherlands, 1985; p 273. (16) Thurman, E. M.; Wershaw, R. L.; Malcolm, R. L.; Pinckney, D. J. Org. Geochem. 1982, 4, 27. (17) Martin, M.; Williams, P. S. In Theoretical Advancement in Chromatography and Related Separation Techniques; Dondi, F., Guiochon, G., Eds.; Kluwer Academic Publishers: Dordrecht, The Netherlands, 1992; p 513. (18) Wahlund, K. G.; Winegarner, H. S.; Caldwell, K. D.; Giddings, J. C. Anal. Chem. 1986, 58, 573. (19) Stevenson, S. G.; Ueno, T.; Preston, K. R. Anal. Chem. 1999, 71, 8. (20) Williams, S. K. R.; Keil, R. G. J. Liq. Chromatogr. Relat. Technol. 1997, 20, 2815. (21) Beckett, R.; Bigelow, J. C.; Jue, Z.; Giddings, J. C. In Aquatic Humic Substances: Influence on Fate and Treatment of Pollutants; Suffet, I. H., MacCarthy, P., Eds.; American Chemical Society: Washington, DC, 1989; Chapter 5. (22) Hasselov, M.; Lyven, B.; Haraldsson, C.; Sirinawin, W. Anal. Chem. 1999, 71, 3497. (23) Lyven, B.; Hasselov, M.; Haraldsson, C.; Turner, D. R. Anal. Chim. Acta 1997, 357, 187. (24) Thang, N. M.; Geckeis, H.; Kim, J. I.; Beck, H. P. Colloids Surf. 2001, 48, 289.
2814
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 13, 2002
(25) Lead, J. R.; Wilkinson, K. J.; Balnois, E.; Cutak, B. J.; Larive, C. K.; Assemi, S.; Beckett, R. Environ. Sci. Technol. 2000, 34, 3508. (26) Zanardi-Lamardo, E.; Clark, C. D.; Zika, R. G. Anal. Chim. Acta 2001, 443, 171. (27) Reschiglian, P.; Melucci, D.; Zattoni, A.; Mallo´, L.; Hansen, M.; Kummerow, A.; Miller, M. Anal. Chem. 2000, 72, 5945. (28) Pos, W. H.; Riemer, D. D.; Zika, R. G. Mar. Chem. 1998, 62, 89. (29) Riemer, D. D.; Milne, P. J.; Zika, R. G.; Pos, W. H. Mar. Chem. 2000, 71, 177. (30) Milne, P. J.; Riemer, D. D.; Zika, R. G.; Brand, L. E. Mar. Chem. 1995, 48, 237. (31) Petasne, R. G.; Zika, R. G. Mar. Chem. 1997, 56, 215. (32) Pos, W. H.; Milne, P. J.; Riemer, D. D.; Zika, R. G. J. Geophys. Res. 1997, 102, 12831. (33) Lee, T. N.; Clarke, M. E.; Williams, E.; Szmant, A. F.; Berger, T. Bull. Mar. Sci. 1994, 54, 621. (34) Ortner, P. B.; Lee, T. N.; Milne, P. J.; Zika, R. G.; Clarke, M. E.; Podesta, G. P.; Swart, P. K.; Tester, P. A.; Atkinson, L. P.; Johnson, W. R. J. Geophys. Res. 1995, 100, 13,595. (35) Sharp, J. H.; Benner, R.; Bennett, L.; Carlson, C. A.; Fitzwater, S. E.; Peltzer, E. T.; Tupas, L. M. Mar. Chem. 1995, 48, 91. (36) Moore, C. A.; Farmer, C. T.; Zika, R. G. J. Geophys. Res. 1993, 98 (2), 2289. (37) Giddings, J. C.; Yang, F. J. F.; Myers, M. N. Science 1976, 193, 1244. (38) Giddings, J. C.; Yang, F. J. F.; Myers, M. N. Anal. Chem. 1976, 48 (8), 1126. (39) Hosse, M.; Wilkinson, K. J. Environ. Sci. Technol. 2001, 35, 4301. (40) Specht, C. H.; Kumke, M. U.; Frimmel, F. H. Water Res. 2000, 34, 4063.
Received for review November 14, 2001. Revised manuscript received March 29, 2002. Accepted April 24, 2002. ES015792R
