Pseudopolarographic Determination of Cd 2+ Complexation in Freshwater JEFFREY J. TSANG,* TIM F. ROZAN, HEILEEN HSU-KIM, † KATHERINE M. MULLAUGH, AND GEORGE W. LUTHER, III* University of Delaware, College of Marine Studies, 700 Pilottown Road, Lewes, Delaware 19958, U.S. Pseudopolarography was used to detect Cd 2+ complexes in samples collected at several locations along the Potomac River in June and September, 2004. Irrespective of site and sampling time, no weak inorganic Cd 2+ species were present. However, up to two stable Cd 2+ -organic complexes were detected at each site. These unknown Cd 2+ complexes were characterized by their half- wave potential (E 1/2 ). The E 1/2 values indicated certain Cd 2+ complexes were common at different sites during each sampling but different complexes were observed in June and September. A Cd 2+ chelate scale, generated from model ligands, was used to estimate the thermodynamic stability constants (K THERM ) of the unknown complexes, which ranged from log K THERM ) 21.5-32.0. Pseudopolarography did not recover all Cd 2+ in the samples. This was partly attributed to highly stable Cd-sulfide species; owing to the presence of acid volatile sulfide at concentrations greater than total dissolved Cd 2+ . These electrochemically inert species may be multinuclear Cd-sulfide clusters and/ or nanoparticles with K THERM values that exceed the detection window of pseudopolarography (log K THERM > 34.4). 1. Introduction Cadmium is regarded as a priority contaminant in water bodies because it is a low level toxicant that has a strong tendency to bioaccumulate in the food web (1). The bio- availability and toxicity of Cd is mainly controlled by its dissolved free metal ion concentration (2). Decreasing the concentration of free metal ion by complexation with organic ligands diminishes toxicity (3). This is generally true for Cd (4); however, a few studies reported increased toxicity of Cd upon complexation as it facilitated Cd 2+ uptake by daphnid (5), phytoplankton, and zooplankton (6, 7). Voltammetric stripping techniques are widely used to study metals in aquatic environments as they offer (sub)- nanomolar detection limits, extremely high sensitivity, and the ability to differentiate between different physicochemical forms (8). Anodic stripping voltammetry (ASV) has been widely used to investigate complexation of Cd 2+ (2, 8-20) as it is able to react directly on the electrode to form an amalgam [Cd(Hg)]. Pseudopolarography has been successfully used to study metal complexation with a variety of ligands in model electrolyte solutions and natural waters (2, 9, 11, 21-29). This technique does not require addition of chemicals that would otherwise change the natural equilibrium and dis- tribution of species; and thus allows the study of the actual metal complexes in solution. Previous studies (21, 24, 25, 30, 31) have fully described the theory of pseudopolarography and its application to metal speciation; a synopsis is given below (see the Supporting Information for more detail). Pseudopolarography involves performing several suc- cessive ASV experiments; each with a different deposition potential (Edep) that is varied more negatively from the free metal reduction potential. The Cd 2+ complexes detected by pseudopolarography involve direct irreversible reduction of the complex at the electrode. This electrochemically destroys complexes by breaking metal-ligand bonds. The metal is subsequently reduced to an amalgam at the electrode surface (eq 1). Upon reoxidation during an anodic scan, the complex does not reform during the analysis as only the metal ion gives an anodic current on reoxidation. Thus, anodic currents are due to the reoxidation of Cd(Hg) to free/labile Cd 2+ and not to any complex. The plot of anodic peak currents against Edep yields sigmoid curves analogous to a dc polarogram, and thus is called a pseudopolarogram (28). These curves reflect the thermodynamic and kinetic properties of the complex (21, 23); and are shifted toward more negative potentials relative to the reduction of the free metal ion (32). The magnitude of the shift is directly related to the thermodynamic stability constant (KTHERM) of the complex (eqs 2 and 3). The more negative the half-wave potential (E1/2) of a sigmoid curve the more stable the complex (33). E 1/2 ) observed reduction potential for the complex; E′1/2 ) formal reduction potential for the divalent metal; n ) number of e - transferred; KTHERM ) stability constant for the metal- ligand complex (eq 3) where {} indicates activities of metal (M) and ligand (L). Thermodynamic stability constants (KTHERM) for unknown ligands bound to Zn (22), Cu (25), and Pb (27) in natural waters have been determined from chelate scales for each metal. These were constructed from plots of known log KTHERM values, for selected metal(II) complexes, against corresponding E1/2 values determined by pseudopo- larography. The detection window of this method is limited by Na + reduction that starts to occur at Edep about -1.7 V (27). Hence information on discrete metal complexes using pseudopolarography is limited to Edep equal to and more positive than -1.7 V. Previous pseudopolarographic studies were unable to completely recover all dissolved Zn (22), Cu (25), and Pb (27) in natural samples. This was attributed to electrochemically inert complexes that were organic (25, 27, 34) and/or inorganic in nature. The latter was postulated to involve highly stable multinuclear metal sulfide clusters and/or nanopar- ticles (27, 35-38). These species have been observed in oxic estuarine (39, 40) and fresh (37) waters for Fe, Cu, and Zn. The low reactivity of metal sulfide clusters has been attributed to their extended structures that exhibit covalent bonding (35, 37, 38); which allow them to persist in oxic waters. * Address correspondence to either author. Phone: +1 (302) 645- 4208 (G.W.L.); +1 (302) 645-4257 (J.J.T.). Fax: +1 (302) 645-4007. E-mail: luther@udel.edu (G.W.L.); jtsang@udel.edu (J.J.T.). † Current address: Duke University, Department of Civil and Environmental Engineering, Box 90287, Durham, NC 27708. ML + 2e - f M(Hg) + L 2- (1) E 1/2 ) E′ 1/2 - (0.0591/n)log K THERM (2) K THERM ) ML/[{M 2+ }{L 2- }] (3) Environ. Sci. Technol. 2006, 40, 5388-5394 5388 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 17, 2006 10.1021/es0525509 CCC: $33.50  2006 American Chemical Society Published on Web 08/05/2006