Environ. Sci. Technol. 2008, 42, 7236–7241
Precipitation and Growth of Zinc Sulfide Nanoparticles in the Presence of Thiol-Containing Natural Organic Ligands BORIS L. T. LAU AND HEILEEN HSU-KIM* Civil and Environmental Engineering, Duke University, Durham, North Carolina 27708
Received May 16, 2008. Accepted August 04, 2008.. Revised manuscript received August 02, 2008
In sulfidic aquatic systems, metal sulfides can control the mobility and bioavailability of trace metal pollutants such as zinc, mercury, and silver. Nanoparticles of ZnS and other metal sulfides are known to exist in oxic and anoxic waters. However, the processes that lead to their persistence in the aquatic environment are relatively unknown. The objective of this study was to evaluate the importance of dissolved natural organics in stabilizing nanoparticulate ZnS that precipitates under environmentally relevant conditions. Precipitation and growth of ZnS particles were investigated in the presence of dissolved humic acid and low-molecular weight organic acids that are prevalent in sediment porewater. Dynamic light scattering was usedtomonitorthehydrodynamicdiameterofparticlesprecipitating in laboratory solutions. Zn speciation was also measured by filtering the ZnS solutions (20 days (8)]. Thus, Zn sulfides are persistent with regard to oxidation under oxic conditions and can exist in surface waters well beyond the original sulfidic source (9). In addition to controlling the fate of Zn in aquatic environments that have a source of excess S(-II), sulfide nanoparticles can also serve as sites for adsorption or coprecipitation of toxic metals such as Hg, Ag, and Cd (10-12). During the initial stages of ZnS(s) mineralization, particle growth is controlled by nucleation, crystallization, and aggregation processes (1). Nanoscale particles tend to have high interfacial surface energies relative to larger particles of the same mineral composition. Under conditions of mineral oversaturation, the surface energy instability causes the nucleated nanoparticles to grow through aggregation of the particles and through crystal growth on individual particle surfaces (13). Thus, for metal sulfide nanoparticles and clusters to persist in natural aquatic environments, the surface of the particles must be modified to overcome interfacial instabilities. Ligand-capped precipitation reactions are utilized to synthesize ZnS nanocrystals for optical applications. These same processes may also be occurring in the natural environment, albeit under less controlled chemical conditions (e.g., pH, temperature, concentrations). During crystallization of ZnS nanomaterials, thiolate-containing organics such as cysteine, thioglycolic acid, and mercaptopropionic acid are used as surfactants to stabilize the particles in suspension (14-16). These thiols are also known to persist in anaerobic porewater of sediments (17-19). The objective of this study was to identify environmental conditions that can lead to relatively stable nanoscale particles during ZnS precipitation. In the aquatic environment, surface association of dissolved organics will be critical for controlling the persistence of ZnS nanoparticles. We performed ZnS precipitation experiments in the presence of natural organic ligands, including cysteine, glutathione, and thioglycolate. These compounds are thiol-containing ligands that persist in anaerobic sediment porewater and surface waters in the nanomolar to micromolar range (17-20). Other low-molecular weight ligands tested in this study included serine, glycolate (both hydroxyl-containing analogues to cysteine and thioglyocolate), and oxalate. These compounds enter natural waters as exudates from biota or by degradation of macromolecular natural organics (e.g., refs 21-23). Dynamic light scattering was utilized to monitor and compare observed growth rates of ZnS particles in the presence of the low-molecular weight ligands and also dissolved humic acid. Our results indicated that organic ligands, including lowmolecular weight thiolates and humic acids, were particularly effective in maintaining ZnS particles at the nanoscale.
Introduction Metal sulfide nanoparticles exist in natural waters as intermediates of mineralization reactions (1). These nanoparticles are small enough to pass through conventional filters * Corresponding author address: 121 Hudson Hall, Durham, NC 27708; telephone: (919) 660-5109; fax: (919) 660-5219; e-mail:
[email protected]. 7236
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 19, 2008
Experimental Section Materials. All chemicals utilized for this research were ACS reagent grade from Fisher Scientific, unless otherwise noted. Filtered (17.8 MΩ-cm) was used to prepare all stock solutions. All glass containers were acid-cleaned by being soaked overnight in 1 N HCl, followed by three rinses with ultrapure water. 10.1021/es801360b CCC: $40.75
2008 American Chemical Society
Published on Web 09/05/2008
Trace-metal grade acids were used for pH adjustments. Ultra high purity nitrogen (N2) was used as needed for purging steps. Zn stock solutions consisted of Zn(NO3)2 · H2O dissolved to a final concentration of 5 mM in 0.1 M HNO3. Sulfide stock solutions were prepared by dissolving crystals of Na2S · 9H2O (rinsed with ultrapure water and dried prior to weighing) in N2-purged ultrapure water. Sulfide stock solutions were stored at 4 °C and utilized within 12 h. A stock solution of the Pahokee peat humic acid standard (International Humic Substances Society) was prepared in a buffer solution containing 4 mM sodium 4-(2-hydroxyethyl)piperazine-1-ethanesulfonate (HEPES) buffer adjusted to pH 7.6. Individual stock solutions of 5.0 mM L-cysteine, 5.0 mM thioglycolic acid, and 5.0 mM L-serine were stored at 4 °C and utilized within 1 week of preparation. Sodium oxalate and glycolic acid stock solutions were stored at room temperature. ZnS Sample Preparation. The Zn sulfide precipitation studies were conducted in an aqueous buffer solution consisting of 0.01 M NaNO3 and 4.0 mM HEPES (pH 7.6). KCl was used instead of NaNO3 in a subset of precipitation experiments. Precipitation experiments with humic acid were conducted with 0.1 M NaNO3. The buffer solutions were filtered using 0.2 µm nylon syringe filters (VWR International). Prior to the initiation of ZnS precipitation, a dissolved organic ligand (humic acid, cysteine, serine, thioglycolate, glycolate, glutathione, or oxalate) was added to the buffer matrix. The precipitation of ZnS was initiated by addition of Zn(NO3)2, followed by Na2S (from their respective stock solutions), to yield final concentrations of 2.0 µM Zn and 2.0 µM sulfide. All Zn sulfide samples were prepared in 50 mL polypropylene tubes. In a subset of the samples, the ionic strength was altered by increasing the NaNO3 concentration from 0.01 to 0.5 M prior to addition of the organic ligand, Zn, and sulfide. All test samples were prepared under ambient laboratory (i.e., oxic) conditions. Sulfide stability under oxic conditions was confirmed by measuring the total sulfide concentration in the Zn sulfide/ ligand mixtures by the Cline method (24). The average recoveries ((1 standard deviation) of sulfide were 113% ((3.5%), 91.3% ((11.2%), and 108% ((3.4%) after storage of the Zn sulfide/ligand mixtures for 2 h, 1 day, and 2 days, respectively. These recoveries each corresponded to an average of three or four samples that included either 2.5 µM cysteine, 0.5 µM GSH, 200 µM serine, 200 µM oxalate, or no organic ligand. ZnS Precipitation and Particle Growth. Particle growth was monitored in the Zn sulfide mixtures by dynamic light scattering (Malvern Zetasizer NS). Immediately after addition of Zn and sulfide to the test mixture, a 1 mL aliquot was dispensed into a polycarbonate disposable 1 cm cuvette and placed in the instrument sample holder. Scattering of the incident light (563 nm wavelength) was measured at 173°. The average hydrodynamic diameter of particles was estimated approximately every 7 min by averaging 15-25 individual 10 s measurements. All particle growth measurements were conducted at 25 °C without mixing. The background scattering rate of the buffer matrix containing dissolved organic ligands (but not Zn or sulfide) was between 30 and 40 kilocounts per second (kcps). Thus, particle size measurements were accepted only for samples that achieved scattering rates greater than 60 kcps (for the same signal attenuation and detection settings). The zeta (ζ) potential of the Zn sulfide particles was estimated from electrophoretic mobility measurements of the particle suspensions (Malvern Zetasizer NS). These suspensions consisted of Zn(NO3)2 and Na2S (20 µM each) in HEPES buffer [0.01 M NaNO3 (pH 7.6)]. Electrophoretic mobility was monitored 10 times over the course of 1 h after the initiation of ZnS precipitation.
X-ray photoelectron spectroscopy (XPS) and transmission electron microscopy (TEM) were utilized to confirm the presence of Zn sulfide particles in the samples. After aging for 1 h, the Zn sulfide/ligand mixtures were filtered with 0.05 µm polycarbonate filters (Whatman). XPS was used to determine the Zn and S content of materials that deposited on the filters. XPS data were compared to spectra of sphalerite (SPI-Chem) and wurtzite powder (Sigma-Aldrich). TEM was used in a subset of the Zn sulfide mixtures to confirm the presence of nanoparticles. Selected area electron diffraction (SAED) and energy dispersive X-ray spectroscopy (EDS) were utilized to assess crystallinity and elemental content in the observed particles. Details of the sample preparation, analysis, and results are reported in the Supporting Information. Zn Speciation by Anodic Stripping Voltammetry. The speciation of Zn(II) was assessed in the Zn sulfide/ligand mixtures by anodic stripping voltammetry (ASV) with a hanging mercury drop electrode (Metrohm VA stand 663). We used ASV because dissolved Zn2+ and other Zn-ligand complexes are detected by ASV, whereas ZnS nanoparticles are inert (i.e., not detected) during ASV analysis (25). The Zn sulfide/ligand solutions were allowed to age for 1 h prior to being placed in the electrode sample cup holder and purged with N2. Additional Zn-ligand control samples were prepared to test the lability of aqueous Zn-ligand complexes during the ASV measurement. These control samples were prepared with dissolved Zn(NO3)2 and organic ligand (cysteine, serine, thioglycolate, glycolate, or oxalate). The controls did not contain sulfide. Zn depositions on the electrode surface were conducted at -1.3 V (vs the Ag/AgCl reference electrode) for 10 s. The concentration of Zn was quantified by stripping in square wave mode from -1.3 to -0.1 V (0.025 V pulse height, 0.20 V/s scan rate) and measuring the resultant current peak area at -1.0 V (corresponding to oxidation of Zn0 to Zn2+). The limit of detection was less than 0.05 µM, corresponding to the 10 s deposition time. The concentration of Zn measured by ASV in the test sample (pH 7.6) is termed “ASV-labile Zn” in this study. A subset of the ZnS/ligand mixtures was filtered to 0.2 µm) from forming during the first 90 min of precipitation (Figure 1). Moreover, serine and glycolate (hydroxylcontaining analogues to cysteine and thioglycolate) did not change or only slightly decreased the rate of growth of ZnS particles (Figure 1). These results indicated that the thiol VOL. 42, NO. 19, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
7237
FIGURE 1. Average diameter of ZnS particles precipitating in the presence of cysteine (1.0 µM), serine (100 µM), thioglycolate (4 µM), glycolate (100 µM), oxalate (100 µM), or peat humic acid (0.56 mg of C/L). Precipitation was initiated by dissolving 2 µM Zn(NO3)2 and 2 µM Na2S in an aqueous buffer containing 4 mM HEPES (pH 7.6) and 0.01 M NaNO3 (with the exception of the humic acid mixture, which contained 0.1 M NaNO3). functional groups on the ligands were interacting with the surfaces of the particles and slowing growth. The time-dependent data shown in Figure 1 were based on intensity-weighted size estimates, which can be biased toward larger particles. Thus, the average diameter resulting from the dynamic light scattering measurements must be interpreted carefully because they do not necessarily reflect the average size based on the number distribution of particles. In this study, we used the intensity-weighted average diameter as a comparative measure of growth rates between the Zn sulfide/ligand mixtures. Filtration of the samples using 0.2 µm filters was used as a secondary confirmation of approximate particle size in the aqueous suspensions. XPS and TEM were utilized to confirm that Zn sulfide particles were forming in the sample mixtures. Zinc sulfide mixtures containing the low-molecular weight ligands (with the same concentrations of mixtures shown in Figure 1) were deposited on a 0.05 µm filter. In all samples, the XPS spectra indicated peaks with binding energies corresponding to zinc as Zn(II) and sulfur as S(-II) (raw data in Supporting Information Figures S2 and S3). The filter material (polycarbonate) produced strong carbon 1s and oxygen 1s peaks. As a result, the carbon and oxygen signals could not be used to differentiate between the sample and filter. The positions of the Zn and S peaks were consistent with XPS spectra of sphalerite and wurtzite reference compounds. The presence of ligands did not result in any discernible shift in Zn and S peak positions. In all samples, the elemental content of the samples was calculated using the relative magnitudes of the Zn 2p and S 2p peaks. The results indicated that the surface of the particles contained a slight excess of sulfur relative to zinc (approximately 0.72 mol of Zn/mol of S), which was consistent with the ZnS(s) references. Electron diffraction of particles observed by TEM indicated that the particles were amorphous or disordered. Moreover, EDS spectra of particles observed by TEM indicated the presence of Zn and S (Supporting Information Figure S5). Zn Speciation by Voltammetry. Characterization of the Zn sulfide samples by XPS, electron diffraction, and EDS involved filtration of samples or TEM images of a small fraction of dried particles. Thus, the data acquired by these analyses did not quantitatively indicate the portion of Zn as Zn sulfides in the samples. Anodic stripping voltammetry (ASV) was used to further confirm the speciation of total Zn(II) in bulk solution by quantifying the fraction of Zn coordinated to sulfide. As shown in previous studies (7, 25), Zn(II) becomes electroactively inert (i.e., not detected by ASV) when dissolved in solution with sulfide. In contrast, 7238
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 19, 2008
FIGURE 2. Total and labile Zn(II) that passed through 0.2 µm filters after being aged for 30 min. Labile Zn was quantified by anodic stripping voltammetry (ASV). The Zn sulfide/ligand mixtures contained 2 µM Zn(NO3)2 and 2 µM Na2S dissolved with an organic ligand: either 1.2 mg of C/L humic acid, 2 µM cysteine (cys), 2 µM serine (ser), or 200 µM oxalate (pH 7.6, 0.1 M KCl). The inset shows ASV-labile Zn(II) in Zn-ligand mixtures using the same ligand concentration. In all Zn-ligand mixtures, Zn(II) was detected by ASV except for the Zn sulfide solution. aqueous Zn(II) complexes with dissolved ligands (including peat humic acid, cysteine, serine, and oxalate) can be detected by voltammetry. ASV-labile Zn(II) and filterable Zn(II) were quantified in the Zn sulfide/ligand mixtures (Figure 2). In control samples containing dissolved Zn(II)-ligand complexes with humic acid, cysteine, serine, or oxalate, between 73.0 and 99.7% of the total Zn was detected as labile during the ASV measurement (Figure 2, inset). In contrast, less than 3% of the Zn was measured in the Zn(II) sulfide samples. From these results, we used ASV to differentiate between Zn sulfides and Zn-organic ligand complexes in the test mixtures. The ZnS/ligand solutions were filtered (
