Effect of Ozone on Algae as Precursors for Trihalomethane and

Aug 9, 2001 - Hee-Jong Son , Hong-Ki Park , Young-Do Hwang , Jong-Moon Jung ... Saif Ur Rehman , Madeha Sadiq , Tariq Mahmood , Jong-In Han...
0 downloads 0 Views 244KB Size
Environ. Sci. Technol. 2001, 35, 3661-3668

Effect of Ozone on Algae as Precursors for Trihalomethane and Haloacetic Acid Production JEANINE D. PLUMMER* AND JAMES K. EDZWALD Department of Civil and Environmental Engineering, 18 Marston Hall, University of Massachusetts, Amherst, Massachusetts 01003

The effect of ozone on the trihalomethane (THM) and haloacetic acid (HAA) formation of two algae species was investigated. Scenedesmus quadricauda (green alga) and Cyclotella sp. (diatom) were cultured under controlled conditions and harvested in the log or late log growth phase. Experiments examined the formation of disinfection byproducts (DBPs) from the algal suspensions with and without preozonation. Preozonation with 1 mg/L increased chloroform formation from Scenedesmus by 17-44%. For Cyclotella, chloroform production increased by 5-26% with 1 mg/L ozone and by 39-109% with 3 mg/L ozone. Chlorinated HAA yields were not significantly increased after 1 mg/L ozone but increased by 38-76% for Cyclotella after 3 mg/L ozone. As compared to other sources of organic matter, algae under bloom conditions may contribute significantly to the DBP precursor pool. However, the majority of the DBP precursors (70%) were attributable to the cellular material, and thus removal of algae cells from a drinking water supply prior to oxidation will substantially reduce algal precursor concentrations.

Introduction It is well-known that chlorination of drinking waters produces trihalomethanes (THMs) and other chlorinated byproducts. Naturally occurring organics, such as humic and fulvic materials, serve as precursors for these byproducts (1). Algae cells and their excreted metabolic products may also contribute to the disinfection byproduct (DBP) precursor pool, producing both THMs and haloacetonitriles upon chlorination. However, THM production levels vary based on the algae species, algal growth phase, and reaction conditions including pH, time, and chlorine dose (2-6). Additionally, algae cells and biomass account for the majority of THM precursors, while extracellular organic matter (EOM) produces only a small fraction (7). These results are significant in that algae cells may be effectively removed by conventional water treatment processes while EOM may not be similarly removed. As compared to humic and fulvic acids, algae can contribute significantly to the DBP precursor pool. Laboratory experiments with four algae species in varying stages of growth showed that maximum chloroform yields from algal cultures exceeded yields from humates (5). However, additional research showed chloroform production from 1 mg/L * Corresponding author present address: 100 Institute Road, Department of Civil and Environmental Engineering, Worcester Polytechnic Institute, Worcester, MA 01609; phone: (508)831-5142; fax: (508)831-5808; e-mail: [email protected]. 10.1021/es0106570 CCC: $20.00 Published on Web 08/09/2001

 2001 American Chemical Society

of the blue-green algae Anabaena was 48 µg/L after a 5-d reaction period, as compared to 80 µg/L for fulvic acid under the same conditions (8). The importance of algae as a precursor material has also been confirmed in natural waters. Correlations have been found between algal activity and THM formation potential (THMFP) concentrations in open surface waters (9). These THMFP concentrations varied throughout the day: highest levels coincided with peak extracellular production time (10). There is no data in the refereed literature on haloacetic acid (HAA) formation from algae or on the effect of ozonation on DBP formation from algae. Such data are useful in light of the Disinfectants and Disinfection Byproduct (D/DBP) Rule, which sets an MCL of 0.080 mg/L for total THMs and 0.060 mg/L for total HAAs (11). Stage 2 of this rule may set more stringent limits by requiring location running annual averages to meet the MCLs rather than averages across the distribution system. This paper presents additional data on THM formation from algae and provides new data on (i) algae as precursors for haloacetic acids, (ii) the effect of ozone on the concentration and composition of disinfection byproducts from algae, and (iii) the importance of algae as DBP precursors. Two algae species were used: Scenedesmus quadricauda and Cyclotella sp. These algae were chosen because they are both dominant surface water algae and can pose difficulties in drinking water treatment. The specific research objectives were (i) to determine DBP formation for the green algae, Scenedesmus, and the diatom, Cyclotella; (ii) to determine the effect of preozonation on DBP formation; and (iii) to determine the relative contribution of algae to the DBP precursor pool as compared to other sources of organic matter. These objectives were part of a larger study that examined ozone and chlorine effects on algae cell properties, EOM, and coagulation and separation of algae (12).

Experimental Section Algal Culturing. Axenic cultures of S. quadricauda (green algae, UTEX Collection No. 76) and Cyclotella sp. (diatom, UTEX Collection No. 1269) were obtained from the Culture Collection of Algae at the University of Texas at Austin (13). Scenedesmus was cultured in the late log growth phase in a continuously mixed, 15-L chemostat. The chemostat was supplied with a constant inflow of synthetic, sterilized algal growth media (see Table 1). A gravity overflow port maintained a constant volume. The chemostat was housed in a 25 °C water bath, supplied with filter sterilized air, and provided with 350 fc (foot-candles) of illumination on a 16-h light/8-h dark cycle. Steady-state conditions were monitored by particle counts (MetOne WGS 260 grab sampler, LB-1010 sensor; MetOne, Inc., Grants Pass, OR) and microscope counts on a daily basis as evidenced by little or no change in growth with respect to time. The cell concentration in the chemostat was maintained at 500 000-600 000 cells/mL, and cell size ranged from 5 to 20 µm. Cyclotella was grown in batch mode in 500-mL Erlenmeyer flasks containing 250 mL of the algal growth media. Cultures received 200 fc of light on a 12-h light cycle and were maintained at 17 °C. Growth was monitored daily by particle counting. Stock cultures were prepared weekly by aseptic transfer of a known number of algal cells from a culture in the log growth phase to a freshly sterilized flask with growth media. Cultures were harvested in the log growth phase (at cell concentrations of 250 000-350 000 cells/mL and cell size of 4-8 µm) for the experiments. VOL. 35, NO. 18, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

3661

TABLE 1. Composition of Algal Growth Media conc in nutrient media components CaCl2‚2H2O MgSO4‚7H2O NaHCO3 Na2SiO3‚9H2O Na2EDTA H3BO3 KNO3 Na2HPO4 FeCl3‚6H2O CuSO4‚5H2O ZnSO4‚7H2O CoCl2‚6H2O MnSO4‚H2O Na2MoO4‚2H2O biotin B12 thiamine

(mg/L)

(µg/L)

Primary Components 36.8 37.0 12.6 28.5 4.36 1.00 131 8.70 3.16 Trace Components 1.00 2.30 1.19 15.2 0.66 0.05 0.05 1.00

As shown in Table 1, no bromide was added to the algal growth media. Therefore, the results presented here apply to low bromide waters, which represent a large fraction of drinking waters. Additionally, only chlorinated DBP data are shown, as concentrations of brominated THM and HAA species were insignificant. Scanning Electron Micrographs (SEMs). Algae samples (100 000 cells/mL) were concentrated by settling, fixed with 2% glutaraldehyde, and washed with 0.025 M sodium phosphate buffer. The fixed and washed cells were then deposited on a cover glass pretreated with 3-aminopropyltriethoxysilane. The cover glasses were placed in 1% OsO4 for 20 min and rinsed. The specimens were dehydrated with successively higher concentrations of ethanol, followed by critical point drying. The dried samples were mounted on aluminum stubs with silver paint and sputter-coated with gold-palladium. The specimens were observed and photographed in a JEOL JSM-5400 scanning electron microscope at 5 or 10 kV. EOM Extraction and Quantification. EOM was extracted from the algae cultures and separated into different apparent molecular weight fractions before and after ozonation. Previously tested methods for EOM extraction were adapted for use here (14). Loosely bound polymers and macromolecules were stripped from the algae cells by shear agitation in a blender for 30 s on high speed. After the blending, the suspensions were scrutinized microscopically with reverse India ink stain, which revealed that the stripping process did not cause any cell damage. The agitated algal culture was filtered through a 1.2-µm retention GF/C filter to remove cells (Whatman International Ltd., Maidstone, England). The filtrate contained the EOM, while the retained matter, resuspended in reagent water to the original suspension volume, contained the cells. The apparent molecular weight distribution of the EOM was determined by dissolved organic carbon (DOC) by ultrafiltration using previously tested techniques (15). The extracted EOM was filtered through Amicon YM and YC series membranes in parallel in 250-mL pressurized, stirred cells (Amicon Inc., Beverly, MA). Pressure was provided by a nitrogen atmosphere at 55 psi, according to the manufacturer’s recommendation. The cells were filled with 120 mL of extracted EOM, stirred, and pressurized. The first 10 mL of filtrate was discarded, and then 30 mL of the filtrate was collected for DOC analysis. Nominal molecular weight limits 3662

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 18, 2001

of 500, 1000, 3000, 10 000 and 30 000 were used. Parallel filtration provided a series of permeates that were analyzed for DOC concentration (Shimadzu TOC-5000 total organic carbon (TOC) analyzer, Shimazdu Corp., Kyoto, Japan). Prior studies (16, 17) have shown that rejection of ultrafiltration membranes is greater for lower molecular weight cutoff membranes. However, two conditions of the experiments in this study minimize mass accumulation at the membrane surface: (i) DOC concentrations were low (5 mg/L residual. Preliminary experiments showed that 15 mg/L (chlorine to TOC ratio of 14) produced the appropriate residual concentration. The chlorine stock concentration was measured by titration in accordance with the DPD Ferrous

FIGURE 1. Scanning electron micrographs of Scenedesmus quadricauda before and after ozonation. (a) No ozone, (b) 1 mg/L ozone, and (c) 3 mg/L ozone.

FIGURE 2. Scanning electron micrographs of Cyclotella sp. before and after ozonation. (a) No ozone, (b) 1 mg/L ozone, and (c) 3 mg/L ozone.

FIGURE 3. Effect of ozone on chloroform formation for 20 000 cells/ mL Scenedesmus (pH 7.0, 20 °C). Titrimetric Method [Method 4500-Cl F (19)], and chlorine residual was measured spectrophotometrically [DPD Colorimetric Method, Method 4500-Cl G (19)]. DBP Quantification. Six HAA species were measured (HAA6: monochloroacetic acid, dichloroacetic acid, trichloroacetic acid, monobromoacetic acid, dibromoacetic acid, and bromochloroacetic acid). Concentrations of chlorinated

species are reported in this study as brominated species were insignificant. After the reaction period, residual chlorine was quenched with sodium sulfite. HAA samples were processed using a modified version of Standard Method 6251B (19). In the modified method, methylation is achieved via incubation in an acidic methanol solution at 50 °C, thus eliminating the diazomethane addition step (20). Total THMs were also measured. Brominated species were negligible; therefore, chloroform concentrations are reported. THM samples were processed with a pentane extraction method similar to Standard Method 6232B, with modifications (21). Concentrations of HAAs and THMs were then determined by gas chromatography with electron capture detection (HP 5890 series II; Hewlett-Packard, Wilmington, DE). Duplicate samples were processed for each experimental condition, and reproducibility was (5%. Average results for chlorinated HAAs and chloroform are presented in the Results section.

Results and Discussion SEMs. SEMs of the algae before and after oxidation demonstrate the effect of the oxidants on cell structure and cell wall features. Figure 1 shows SEMs for Scenedesmus before and after ozonation. The untreated cell shows a colony of four cells with a loosely fitting reticulate layer. This net-like layer is held off the coenobium by tubular propping spikelets VOL. 35, NO. 18, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

3663

FIGURE 4. Effect of ozone on chloroform formation for 20 000 cells/ mL Cyclotella (pH 7.0, 20 °C).

FIGURE 6. Effect of ozone on chlorinated HAA formation for 20 000 cells/mL Cyclotella (pH 7.0, 20 °C).

FIGURE 5. Effect of ozone on chlorinated HAA formation for 20 000 cells/mL Scenedesmus (pH 7.0, 20 °C).

TABLE 2. Dissolved Organic Carbon Concentration of Algal Suspensions (20 000 cells/mL) before and after Ozonation DOC (mg/L) pretreatment

Scenedesmus

Cyclotella

before ozonation after 1 mg/L ozone after 3 mg/L ozone

0.22 0.87 1.13

0.33 0.73 0.89

TABLE 3. Molecular Weight Fractionation of EOM before and after Ozonation (mg/L as DOC (% of total DOC)) 10-3K

10K

and rosettes (22). Four spines are readily apparent. With a 1 mg/L ozone dose, the cell wall appears more loosely folded, but there is no evidence of lysing. An ozone dose of 3 mg/L resulted in severe cell wall alterations. The reticulate layer is detached from the spikelets, and both the spikelets and trilaminar sheath are readily visible. For Cyclotella (Figure 2), untreated cells show a silicified cell wall and radial symmetry. The cell wall, or frustule, consists of two lid-like valves, one fitting inside the other. The overlapping area of the valves is connected by bands 3664

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 18, 2001

FIGURE 7. Total (a) chloroform and (b) chlorinated haloacetic acid formation potential for EOM and cells extracted from 20 000 cells/ mL Cyclotella with and without preozonation (7 day reaction time, pH 7.0, 20 °C). that constitute the girdle. The cell wall and girdle show no apparent damage. After 1 mg/L ozone, the majority of the cells show no damage. For a few cells, there is minor damage to the girdle. After 3 mg/L ozone, a significant number of cells were still not damaged. Some cells showed warping or penetration of the girdle; however, perforation of the cell wall was not observed. DOC Concentration. The DOC concentration of the bulk algal suspensions was measured before and after ozonation. Results are shown in Table 2. For Scenedesmus at a concentration of 20 000 cells/mL, a l mg/L ozone dose resulted in an increase in DOC from 0.22 to 0.87 mg/L. A 3 mg/L ozone dose further increased the concentration of carbon in solution, as indicated by a DOC concentration of 1.13 mg/L. For Cyclotella, similar increases in DOC were observed after ozonation. As was demonstrated by the SEMs for both algae, extensive cell destruction did not occur when the algae were pretreated with 1 mg/L ozone. It is hypothesized that the increase in DOC with this ozone dose was a result of increased liberation of extracellular organic matter. With 3 mg/L ozone,

FIGURE 8. Comparison of chloroform production from algae, humic acids, and fulvic acids.

TABLE 4. Speciation of Chlorinated Haloacetic Acids from Cyclotella before and after Ozonationa (µg/L (% of total HAAs)) ozone dose (mg/L)

time (d)

DCAA

TCAA

total chlorinated HAAs

0 0 0 0 0 1 1 1 1 1 3 3 3 3

1 2 3 5 7 1 2 3 4 7 1 3 5 7

19 (49) 34 (51) 43 (53) 56 (58) 61 (56) 24 (50) 38 (54) 47 (56) 63 (61) 71 (60) 40 (57) 80 (66) 96 (69) 103 (69)

20 (51) 33 (49) 38 (47) 41 (42) 47 (44) 24 (50) 32 (46) 36 (44) 40 (39) 47 (40) 30 (43) 42 (35) 43 (31) 47 (31)

39 66 81 97 108 47 70 84 103 118 69 122 139 150

a DCAA, dichloroacetic acid; TCAA, trichloroacetic acid. Yields (percentages) may not add up to total HAAs (100%) due to rounding errors.

lysing may also have contributed to the increased DOC concentration, in particular for Scenedesmus. Apparent Molecular Weight Fractionation of EOM. Table 3 shows the apparent molecular weight fractionation of EOM extracted from Scenedesmus and Cyclotella, with and without preozonation. To have high enough DOC concentrations after fractionation, a high cell concentration was used for these experiments (approximately 500 000 cells/mL for Scenedesmus and 350 000 cells/mL for Cyclotella). Before ozone, 20% of the EOM from Scenedesmus was found in the >10K molecular weight fraction, 24% was in the 10-3K fraction, and 57% was in the 10K and the 10-3K ranges resulted in an increase in the fraction of EOM found in the