Environ. Sci. Technol. 2005, 39, 1523-1531
Occurrence and Solid-Liquid Partition of Sulfonated Naphthalene-Formaldehyde Condensates in the Aquatic Environment F R A N K T . L A N G E , * ,† MICHAEL MERKLINGER,† MICHAEL WENZ,† HEINZ-J. BRAUCH,† MARKUS LEHMANN,‡ AND ISTVAN PINTER‡ DVGW-Technologiezentrum Wasser (TZW), Karlsruher Strasse 84, D-76139 Karlsruhe, Germany, and State Institute for Environmental Protection Baden-Wu ¨ rttemberg, Griesbachstrasse 1, D-76185 Karlsruhe, Germany
Sulfonated naphthalene-formaldehyde condensates (SNFC) are high production volume chemicals used in a variety of applications, for example, as concrete plasticizers, tanning agents, or dye dispersants. They enter the aquatic environment primarily by the wastewater path. The occurrence and fate of the monomers, which are different isomers of mono- and disulfonated naphthalene, was intensively investigated in previous studies. However, the environmental fate of the persistent higher molecular SNFC is so far widely unknown. This paper describes an ultrasonic extraction under alkaline conditions, followed by ion-pair HPLC with fluorescence detection for the analysis of SNFC oligomers from solid environmental matrixes such as sewage sludge, suspended solids, and river sediments. Limits of quantification of about 0.1 mg kg-1 d.m. were well below the measured concentrations in environmental samples. SNFC were adsorbed to suspended solids and river sediments in three major German rivers (Rhine, Neckar, and Danube) in concentrations typically up to several mg kg-1 d.m. A total content of about 4 g kg-1 d.m. was measured in a sewage sludge of a municipal wastewater treatment plant, which receives wastewater from a textile dyeing plant. Furthermore, the first quantitative field data on the partition of SNFC and their monomers between the aqueous phase and solid environmental compartments are presented. Solid-liquid partition coefficients (Kd) of oligomers with a chain-length ranging from three to six naphthalenesulfonate units were derived from the analysis of corresponding wastewater and sewage sludge samples and from suspended solids and river water samples, respectively. Determined Kd values were in the range from 102 to 104 L kg-1.
Introduction Sulfonated naphthalene-formaldehyde condensates (SNFC, Figure 1) are used worldwide as dispersants in manifold * Corresponding author phone: ++49 721 9678 157; fax: ++49 721 9678 104; e-mail:
[email protected]. † DVGW-Technologiezentrum Wasser (TZW). ‡ State Institute for Environmental Protection BadenWu ¨ rttemberg. 10.1021/es040081p CCC: $30.25 Published on Web 02/05/2005
2005 American Chemical Society
FIGURE 1. Condensation reaction of 2-naphthalenesulfonate with formaldehyde leading to SNFC oligomers of different chain-length; n represents the number of naphthalenesulfonate units within the molecule. applications. They play an important role as concrete plasticizers, as synthetic tanning agents, and as dispersants for dyes. The annual worldwide SNFC consumption of concrete plasticizers and dye dispersants is estimated at 150 000 t each (1). The composition of the active SNFC ingredients can vary to some extent, depending on the production process and on the type of the intended application. Typical products contain about 65% oligomeric and polymeric condensates of 2-naphthalenesulfonate (2NS) with formaldehyde, 10% mono-, and 25% disulfonated monomers as well as traces of other byproducts such as dinaphthyl sulfones and naphthalene (2-4). The acute toxicity of technical SNFC mixtures is low. LD50 values for rats were reported to be higher than 2000 mg kg-1; typical LC50 values for fish were in the range of about 1000-1600 mg L-1 (see ref 5 and material data sheets cited therein). Data on chronic toxicity are not available. However, SNFC may remobilize other pollutants in the environment due to their polyelectrolyte character. SNFC molecules are multiple negatively charged oligomers and polymers (6, 7). In technical processes, they adsorb onto the surfaces of solid particles, for example, of cement grains or dye particles, and avoid their coagulation due to electrostatic repulsion. The tendency to adsorb onto solids also suggests an accumulation of SNFC in sewage sludge, suspended solids, and river sediments. During the past decade, analytical methods for trace determination of monomeric naphthalenesulfonates (8-11) and oligomeric SNFC (7, 12) were developed and applied to aqueous environmental samples. These methods allow for the detection of SNFC residuals in the ng L-1 to µg L-1 range. Ruckstuhl (2, 13) applied an ion-pair chromatographic method with fluorescence detection (7) to wastewater and groundwater samples of two tunnel construction sites in Switzerland, where SNFC were applied as concrete plasticizers. These investigations demonstrated that the SNFC emissions from construction sites into the aquatic environment are rather low. Even under worst-case conditions, when cement suspensions including SNFC were injected into a groundwater aquifer for stabilization of the gravel, a mass flux analysis revealed that only about 5% of the applied SNFC, predominantly monomers, reached the corresponding groundwater. This leaching behavior is confirmed by results of laboratory leaching tests with a hardened cement paste (14). According to these experiments, only SNFC monomers migrated from the cement specimen into the aqueous phase, whereas oligomers and polymers remained immobilized in the cement matrix. The field study at the tunnel construction sites (2) and also a tracer experiment with a technical SNFC product in a gravel aquifer of the Upper Rhine valley, Germany (15), demonstrated the low mobility of SNFC with three or more 2-naphthalenesulfonate units in the molecule, which can be explained by retardation due to adsorption to the aquifer material. In contrast to the low emissions from the modern concrete technology, larger discharges of SNFC into the aquatic environment are expected from other sources, especially from VOL. 39, NO. 6, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
1523
FIGURE 2. Most frequently sampled sites on the Rhine, Neckar, Danube, and Kocher rivers in the state of Baden-Wu1 rttemberg, Germany.
wastewater. Industrial effluents of SNFC manufacturing plants (2, 3, 7), of paper manufacturing plants (3, 7), as well as tannery wastewater (3) have been identified as such sources. Textile dyeing wastewater was also assumed to be an important source of SNFC (16), but was not yet confirmed by any measurement. Naphthalenemonosulfonates and most of the monomeric naphthalenedisulfonates of technical SNFC products are typically biodegraded during sewage treatment (16, 17). Therefore, sewage treatment plant effluents mainly contain the persistent naphthalenedisulfonate isomers and low molecular SNFC oligomers. The concentrations measured in river water samples reflect the emissions by the sewage treatment plant effluents. In general, persistent naphthalenedisulfonates (for details, see ref 18 and references cited therein) and part of the low molecular SNFC fractions up to n ) 5 (7) have been identified in the water body of European rivers. However, the environmental fate of the higher molecular SNFC is so far unknown. Investigations on their adsorption potential to solid environmental matrixes were hindered due to a lack of an appropriate analytical method. In this paper, we describe a new extraction method for SNFC from solid matrixes, which can be combined with an existing high performance liquid chromatographic (HPLC) method for quantitation. Furthermore, we present quantitative field data on the partition of the SNFC oligomers between the water body and the corresponding suspended solids and sediments, which was generated by application of this method in a monitoring program in southwest Germany (Baden-Wu ¨ rttemberg). 1524
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 6, 2005
Experimental Section Materials. HPLC grade water was prepared by a Milli-QSystem (Millipore, Molsheim, France). Hydrochloric acid (HCl) and sodium hydroxide (NaOH), for adjusting aqueous samples to pH 6.5 prior to extraction and for alkaline extraction of SNFC from solid samples, were of p.a. quality (Merck, Darmstadt, Germany). Tetrabutylammonium bromide (TBABr, >98%) was purchased from Fluka (Buchs, Switzerland), and tetrabutylammonium hydrogen sulfate (TBAHSO4, >99.8%) was from Mallinckrodt-Baker (Phillipsburg, NJ). Di-sodium hydrogen phosphate dihydrate (Na2HPO4‚2H2O, >99.5%) was obtained from Merck. The sodium salts of the monomeric naphthalenesulfonates, with a purity of >99%, were purchased from Chemos (Regenstauf, Germany). Standards of SNFC homologues with chain-lengths n ) 2-7 were isolated in our laboratory by semipreparative HPLC from a synthetic tanning agent (7). The internal standard used for the HPLC analysis, p-aminodiphenylamine4-sulfonate (97%), was kindly donated by the Bayer AG (Leverkusen, Germany). Sampling and Preservation. In Figure 2, a map is given showing the most frequently sampled sites mentioned in this work. Water Samples and Sewage Sludge. River water samples were taken at selected sampling sites in Baden-Wu ¨ rttemberg, Germany, between April and December 2000. Except for the samples from the monitoring station at Worms on the Rhine, all other samples were collected as grab samples. For measurements along the Rhine river, samples were taken by
TABLE 1. Average Values of Flow and Additional Water Quality Parameters at the Sampling Sites on the Rhine, Neckar, Danube, and Kocher Rivers during the Sampling Period river Rhine
sampling site
conductivity flow in m3 s-1 a in µS cm-1
remarks
Iffezheim, km 333.8 Mannheim, km 424.7 Worms (line 1), km 443.3
representative site representative site cross profile site on the left side of the Rhine river Worms (line 2-4), km 443.3 composite sample of cross profile from the center to the right side of the Rhine river Neckar Mannheim-Feudenheim, km 8.0 representative site Danube Ulm-Bo¨ fingen, km 2582.5 representative site Kocher (Weisser Kocher), km 160.8 reference site, upstream of a textile dyeing plant Hu¨ ttlingen, km 152.4 reference site, downstream of a textile dyeing plant
chloride boron DOCb AOSc in mg L-1 in mg L-1 in g L-1 in µg L-1
1260d 1290e 1410
447 420 601
40.3 32.5 72.3
0.03 0.04 no data
1.8 1.9 2.4j
43 48 no data
1410
524
51.9
no data
2.4j
no data
149f 124g 0.49h
723 505 753
45.1 17.8 48.5
0.15 0.03 0.02
3.4 2.4 5.5
84 66 73
2.24i
813
43.2
0.34
8.3
288
a Long-term mean runoff MQ. b DOC: dissolved organic carbon. c AOS: adsorbable organic sulfur. d Value for site Maxau (km 362). e Value for site Speyer (km 401). f Value for site Rockenau (km 61) × 1.1. g Value for site Neu-Ulm/Bad Held (km 2587). h MQ (1987-2003). i MQ (1995-2003). j Composite sample of the total cross profile (lines 1-4).
boat in the period November 13-17, 2000. At each sampling site of this campaign, a composite sample of three individual samples over the cross section of the Rhine river was collected by means of a built-in sampling device on the research boat “Max Honsell”. This device includes a stainless steel intake pipe, which allows one to draw the water samples at a depth of 90 cm. The volume ratio of the three individual samples was 1:2:1 (left side:middle:right side of the river) to get a representative sample. At the monitoring station at Worms, there are four permanent intakes across the Rhine river, on the left (line 1), left-middle (line 2), middle-right (line 3), and the right side (line 4). At these positions, water was permanently delivered into the station. From these partial streams, every 10 min an aliquot was automatically sampled and combined to a 1 day composite sample. The 1 day composite samples were manually mixed to a 14 days composite sample. Additional samples were taken at other sampling sites of a regular surface water monitoring program (Table 1). Influent, effluent, and sewage sludge were sampled in a municipal wastewater treatment plant (WWTP). This WWTP is influenced by textile dyeing wastewater. The water and sludge samples were filled into amber glass bottles (2 L) with a glass stopper. The bottles contained 200 mg of sodium azide for stabilization. Samples were transported in cooling boxes to the lab, stored at 4 °C, and extracted within 2 weeks. Suspended Solids and River Sediments. Suspended solids were isolated from the rivers by a flow-through centrifuge (Padberg Z61, flow through, 900 L h-1; revolution, 17 000 min-1; sampling time, 4-6 h). Sediment samples were taken from the rivers by a Van Veen type sediment grab sampler. Via this type of sampling, mainly the freshly deposited upper sediment layer was collected. Suspended solids and sediment samples were filled into 150 mL or 1 L polyethylene bottles, transported in cooling boxes to the lab, and stored frozen at -20 °C prior to freeze-drying. Analysis of Naphthalenesulfonate Monomers and SNFC from Aqueous Samples. Analysis of naphthalenesulfonates and SNFC was performed by ion-pair reversed-phase liquid chromatography using a model 1100 HPLC system from Agilent (Waldbronn, Germany) equipped with a quarternary solvent pump, a temperature controlled column compartment, a UV diode-array detector (DAD) for qualitative confirmation, and a fluorescence detector (FLD) for quantitative determination. Monomeric naphthalenesulfonates were analyzed independently from the SNFC according to Lange et al. (19) from 50 mL aqueous samples with an automated on-line ion-pair extraction/ion-pair chromatography method on RP-C18 solid-
phase material, applying a water-methanol gradient elution and the wavelength combination λEx ) 230 nm/λEm ) 340 nm for fluorescence detection. SNFC were analyzed according to Wolf et al. (7) with slight modifications. Oligomers were enriched by off-line extraction from 500 mL aqueous samples by ion-pair-extraction on 1 g of RP-C18 material. Instead of elution with 10 mL of acetonitrile containing 1 mmol L-1 TBABr as proposed in (7), the oligomers were eluted with 5 mL of acetonitrile, followed by 5 mL of methanol, both containing 1 mmol L-1 TBABr. This modification improved the reproducibility of the internal standard recovery. HPLC separation was carried out as described previously (7) on a RP-C18 column (Hypersil ODS C18, 5 µm, 250 × 4 mm i.d., Agilent) applying a binary gradient with acetonitrile as organic modifier in the mobile phase and fluorescence detection at λEx ) 230 nm/λEm ) 360 nm. Linear calibration functions (naphthalenesulfonate monomers, 0.02-0.30, 0.30-1.0, and 1.0-5.0 µg L-1; SNFC, 0.020.10 and 0.10-2.0 µg L-1) were set up with spiked tap-water samples that were analyzed using the entire method including sample extraction. Highly contaminated samples were diluted with pure tap water. Limits of quantification (LOQ) were statistically derived from the calibration functions in the lowest calibration range and are comprised in Table 2. Blank values were in the range of about 3 ( 2 ng L-1 for NS and NDS and 10 ( 5 ng L-1 for SNFC oligomers. Based on the LOQ derived from the calibration functions and taking into account the blanks, the working LOQ was defined at 20 ng L-1. Further details of the two HPLC methods for the determination of naphthalenesulfonate monomers and SNFC were published previously (7, 19). Analysis of SNFC from Sewage Sludge, Suspended Solids, and River Sediments. Sewage sludge samples were centrifuged for 30 min at 8500g. The precipitate was frozen to -18 °C and lyophilized (freeze-dryer model Beta 2-16, Christ, Osterode, Germany). Suspended solids and sediment samples were lyophilized accordingly. Lyophilized sewage sludge and suspended solids samples were manually crushed in a porcelain mortar. The dry river sediment samples were ground in an agate planetary ball mill and sieved with stainless steel sieves. The fraction 0.1 mg kg-1 were reported. Most of the concentrations found were well above this value (see Results and Discussion). The precision of the overall method was investigated by five replicate analyses of a native, homogenized Neckar river suspended solids sample. The results are given in Table 3. A good precision of the entire analytical method was achieved for SNFC with n ) 3-7 providing relative standard deviations (RSD) between 5% and 10%. A lower precision with a RSD of 30% was found for n ) 2, which is reasonable because the native concentration of these smallest condensates in the sample analyzed was very close to the LOQ of 0.05 mg kg-1. The RSD for the sum of all detected SNFC with n > 7 was about 60%. According to the finding of a nearly constant mass-related response for SNFC with n ) 4-7 in HPLC with fluorescence detection (7), these longer chain SNFC with n > 7 were quantified using the average response obtained for the shorter chain SNFC. The comparatively low precision for the fraction n > 7 may result from an incomplete extraction from the sample. These uncertainties might also originate from sorption of the large hydrophobic ion-pairs during workup of the extracts, for example, during filtration. Therefore, presented concentrations for SNFC with n > 7 have to be considered as semiquantitative. Occurrence, Solid-Liquid Partition, and Elimination of Naphthalenesulfonates and SNFC in Sewage Treatment. Occurrence and distribution of naphthalenemonosulfonates (NS), naphthalenedisulfonates (NDS), and SNFC were investigated in a municipal wastewater treatment plant (WWTP) on the upper stretch of the Kocher river in southwest Germany. This WWTP is strongly influenced by pretreated wastewater of a textile dyeing plant. NS, NDS, and SNFC concentrations in the pretreated textile dyeing wastewater (i.e., a partial stream of the total WWTP influent), in the WWTP effluent, and in the sewage sludge are given in Figure 3 and
(March 5-12, 2001). The analyzed sewage sludge sample was a composite sample of 3 days (March 6-9, 2001). As shown in Table 4 and Figure 3, monomers were incompletely removed during the wastewater treatment. The monomers were not significantly adsorbed to sewage sludge. Moreover, their concentration ratios in the aqueous phase strongly changed during the wastewater treatment process, while the isomer pattern of the SNFC oligomers remained constant. The incomplete removal of NS and NDS can be explained in terms of biodegradation. The monosulfonates, 1- and 2-naphthalenesulfonate (1-NS and 2-NS), and the disulfonates, naphthalene-1,6- and -2,6-disulfonate (1,6-NDS and 2,6-NDS), were well eliminated. Lower eliminations were observed for 1,5-, 1,7-, and 2,7-naphthalenedisulfonates (1,5NDS, 1,7-NDS, and 2,7-NDS). The observed differences between the biodegradability of individual NS and NDS isomers are consistent with earlier work on this topic (see, e.g., refs 16, 18-21 and literature cited therein). SNFC were also incompletely removed from the wastewater during the mechanical biological wastewater treatment. The total SNFC concentration in the pretreated textile wastewater was 1.35 mg L-1 as compared to 0.53 mg L-1 in the WWTP effluent. The unaltered isomeric pattern within each group of SNFC homologues is a strong indication for their microbial persistence. This finding underlines the results obtained in field studies and aerobic laboratory degradation experiments by Ruckstuhl et al. (20). In lab-scale experiments, no primary degradation of SNFC occurred within 195 d when the batches were incubated with sludge from two WWTPs receiving wastewater from textile finishing industry (20). As shown in Figure 3, the removal of SNFC is due to accumulation in sewage sludge. The accumulation tendency of SNFC increases with increasing chain-length. Accumulation of SNFC in solid environmental matter was predicted already in an earlier paper (7), investigated by laboratory experiments (13), and is confirmed in this work by direct extraction. The sewage sludge mainly contained SNFC oligomers with n > 2. Like the monomers, condensates with n ) 2 predominantly remained in the aqueous phase. With 4.3 g kg-1 d.m., the total SNFC concentration in sludge is remarkably high for an environmental contaminant. The
FIGURE 3. Ion-pair chromatograms of different samples from a municipal wastewater treatment plant receiving textile dyeing wastewater and of the Kocher river. Table 4. Samples of the textile dyeing plant effluent and the municipal WWTP effluent were composite samples of 7 days
TABLE 4. Kd Values Derived from the Concentrations of SNFC and Their Monomers in a Partial Influent, in the Effluent, and in the Sewage Sludge of the Municipal WWTP on the Upper Kocher River, and in the Receiving Kocher River at Hu1 ttlingen in Southwest Germany, Downstream of the WWTP; the Partial Influent Is a Pretreated Textile Dyeing Wastewater; NS ) Naphthalenesulfonate, NDS ) Naphthalenedisulfonate partial influent March 512, 2001 µg L-1
effluent March 512, 2001 µg L-1
1-NS 2-NS 1,5-NDS 1,6-NDS 1,7-NDS 2,6-NDS 2,7-NDS
250 1700 98 450 230 53 160
3.9 25 35 22 92 0.84 57
n)1 n)2 n)3 n)4 n)5 n)6 n)7 n>7
2900 460 340 240 150 86 74
