Environ. Sci. Technol. 2003, 37, 3845-3852
Determination of Pb Complexation in Oxic and Sulfidic Waters Using Pseudovoltammetry TIM F. ROZAN,† G E O R G E W . L U T H E R , I I I , * ,† DOUG RIDGE,‡ AND SCOTT ROBINSON‡ College of Marine Studies, University of Delaware, 700 Pilottown Road, Lewes, Delaware 19958, and Department of Chemistry and Biochemistry, University of Delaware, Newark, Delaware 19716
Pseudovoltammetry was used to evaluate the actual Pb complexation occurring in natural water samples of varying oxygen and sulfide concentration. In pseudovoltammetry, the potential at which metal-ligand complexes are broken up to form the metal amalgam is used to determine the complexes’ thermodynamic stability constants (KTHERM; corrected for metal and ligand side reaction coefficients) via the Nernst expression. This methodology removes the need for any metal additions and for subsequent modeling using fitting criteria, which provide only conditional stability constant data (KCOND). Using known organic ligands, a chelate scale ranging from log KTHERM ) 4 to log KTHERM ) 20 was developed as a template for comparison with samples collected from two stations of different salinities and at several depths in the Chesapeake Bay. These samples were observed to contain up to five different ligand compounds of unknown structural composition (log KTHERM > 8) with the strongest ligand fraction exceeding log KTHERM > 39 (the maximum observable thermodynamic stability constant due to the reduction of Na+). One possible explanation for the observed complexation is the existence of lead sulfide clusters. This was supported by laboratory experiments using electrochemistry and ICR-FTMS, which confirmed the formation of electrochemically inert multinuclear clusters with high stability constants (e.g., M3S3, log KTHERM ) 62.9). However, in all field samples, (sub)nanomolar levels of acid-leachable sulfide were recovered at pH 5.0-6.2, which could be attributed to dissociation of lead sulfide complexes with moderate acidity. Recovery of sulfide increased from 90%.
Discussion The data collected from the Chesapeake Bay samples provided several interesting results. First, the data showed a gradient in the type of complexation, which included inorganic sulfides in the bottom waters to a range of weak to strong organic ligands in mid-depth and surface waters. In the bottom waters of the mid-bay site B, the presence of free sulfide (>15 µM) and the lack of any measurable ligands using pseudovoltammetry (zero recovery) indicated that the Pb speciation was dominated by lead sulfide cluster formation with stoichiometries equal to or greater than Pb3S3. Several other experiments were conducted to support this assignment VOL. 37, NO. 17, 2003 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
3849
FIGURE 5. Pseudovoltammograms of filtered (0.2 µm) water samples collected from the upper Chesapeake Bay (site A). (Conditions: 30-min deposition time, SWASV-HDME, 200 mV s-1 scan speed).
FIGURE 6. Pseudovoltammograms of filtered (0.2 µm) water samples collected from the mid-Chesapeake Bay (site B). (Conditions: 30min deposition time, SWASV-HDME, 200 mV s-1 scan speed). of Pb complexation. First, acid titrations were conducted on the samples to quantify sulfide pools based on pH (13). After purging free sulfide at pH 6.7 for several minutes, the results showed that the recovered pool of reduced sulfide in the range of pH 5.0-6.2 exceeded the total dissolved Pb concentration. When corrected for S8, polysulfides, and FeS (39, 40), the concentration of sulfide in this fraction was found to be 150-180% greater than the Pb concentration, indicating other metals were also bound to sulfide. Second, a 1 nM Pb2+ aliquot was added to unpurged bottom water samples, then was allowed to equilibrate under argon, and was acid titrated to determine changes in the bound sulfide pools. The resulting acid titration recovered a 1 nM increase in the sulfide released in the pH 5.0-6.2 range, which corresponded to newly formed PbS. This result was not surprising but confirmed that free sulfide was the dominant excess ligand in the bottom waters. If a strong organic ligand were present, it was unable to react with the added Pb2+ before sulfide could. In contrast, the Pb recoveries from pseudovoltammetry increased to >70% in the low oxygen or suboxic zone (3.05.6 µM O2) and to 90% in the oxygenated surface waters of the high salinity site B. These observations follow the sulfide gradient and show that the mid-depth waters are the transition zone from inorganically dominated complexes to organically dominated complexes. In all surface waters, Pb complexation was dominated by organic ligands. In the surface water obtained from the midbay site B, high Pb recoveries (>92%) from a single compound 3850
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 37, NO. 17, 2003
type (C4) were measured by pseudovoltammetry. A correspondingly low recovery of sulfide from acid titrations (at pH >4.0) suggests metal-sulfide compounds are only a minor component of the metal complexation. In the upper bay (site A), a similar type of strong binding compound was also found, again with low acid-labile sulfide concentrations. However, the lower Pb recoveries (58-66%) at site A based on pseudovoltammetry analysis suggest additional strong Pb organic compound(s) may have also been present in the surface waters of the upper bay. This result is similar to the findings for Cu from Vineyard Sound, MA, where at least two strong Cu chelators were observed (23). The complex measured in that study (representing 30-50% of the Cu) had a thermodynamic stability constant of log K ) 36.7), whereas the other(s) was electrochemically inert up to a -1.7 V deposition (log KTHERM > 39.1; 23). These results showed that the Pb complexation in the surface waters of the Chesapeake Bay is dominated by two or more strong compound types (inert compounds and C4) and not by two ligand classes as normally indicated in the literature (1). Furthermore, the measured thermodynamic stability constants for compound C3 in deeper suboxic waters was found to be significantly higher than those determined for Pb-thiol complexes, including the 1:2 complexes in Table 1. Our data question whether mono-thiol compounds are important complexing ligands for trace metals, as recently suggested by some researchers (8, 52, 53), because they are obviously not as strong as compounds C3, C4, and the inert complexes (either inorganic sulfide metal clusters and the organic complexes) found in these waters of site A. In the oxygenated bottom waters of the upper-bay sample location, three compounds (C1-C3) were observed containing over 90% of the total dissolved Pb. Interestingly, this was the only sample that contained three distinct compounds, including the weakest bound Pb compound (C1). These organic compounds were weaker ligands than those found in the surface waters, which may be attributed to either different extracellular chelates (from different organisms) or the breakdown of chelates from organic decomposition. In the mid-bay location, suboxic water samples (where O2 was 70% of the total dissolved Pb. In this case, the decrease in Pb recovery reflects the increased presence of lead sulfide clusters, which is seen by increases in the acid-leachable sulfide recoveries. To evaluate the effect of metal titrations on natural samples, we decided to conduct a series of metal addition experiments to complex any excess ligand(s) that may be present. At the mid-bay sampling location B, 1 nM Pb2+ aliquots were added to filtered subsamples that had been collected from the surface-oxygenated and mid-depth suboxic zones. Prior to Pb2+ additions, the samples were analyzed by pseudovoltammetry. Pb2+ was then added, and the samples were allowed to equilibrate for 24 h. After equilibration, the samples were reanalyzed with pseudovoltammetry, and pre- and post-metal addition pseudovoltammograms were compared (Figure 7). After a 1 nM Pb2+ addition, four compound types (C1-C4) were observed in the mid-bay surface water sample instead of the original compound, C4. However, the presence and concentration of each ligand compound was dependent on the total Pb2+ added (Figure 8). As each type of ligand became saturated with Pb, weaker binding complexes formed, until the Pb had finally formed a complex with the weakest binding Pb compound observed in any waters studied (C1; log KTHERM ≈ 8). After a second 1 nM Pb2+ addition, no new complexes were formed; however, the concentration of Pb that complexed with the C1 compound increased by a corresponding 1 nM. These increases are similar to increases in the ligand classes during Pb titrations in other complexation studies
FIGURE 7. Pseudovoltammograms of filtered (0.2 µm) water samples collected from the mid-Chesapeake Bay (site B) surface water before and after a 1 nM Pb addition (Conditions: 30-min deposition time, SWASV-HDME, 200 mV s-1 scan speed). Postaddition pseudovoltammetry was conducted following a 24-h equilibration period.
FIGURE 8. Pb recovered (complexation capacities) from the different compound types (C1, C2, C3, and C4) found in Chesapeake Bay (site B) surface and mid-depth waters before and after Pb additions. Post-addition pseudovoltammetry was conducted following a 24-h equilibration period. (14, 54). In Kozelka and Bruland (14) and Muller (54), a CLE-CSV methodology determined two Pb ligand classes in estuarine waters. The observed conditional stability constants of the ligand classes ranged from log K of 10.8-12.2 for the strong ligands to log K of 8.8-10.2 for the weak ligands. In both of these studies, Pb titrations were used to determine total ligand concentrations as well as conditional stability constants. Interestingly, the strong class of ligands was saturated with
