- Email: [email protected]
Characterization, origin and aggregation behavior of colloids in eutrophic shallow lake
Water Research 142 (2018) 176e186
Contents lists available at ScienceDirect
Water Research journal homepage: www.elsevier.com/locate/watres
Characterization, origin and aggregation behavior of colloids in eutrophic shallow lake Huacheng Xu a, *, Mengwen Xu b, Yani Li a, c, Xin Liu b, Laodong Guo d, Helong Jiang a a
State Key Laboratory of Lake Science and Environment, Nanjing Institute of Geography and Limnology, Chinese Academy of Sciences, Nanjing 210008, China b College of Biology and the Environment, Nanjing Forestry University, Nanjing 210037, China c University of Chinese Academy of Sciences, Beijing 100049, China d School of Freshwater Sciences, University of Wisconsin-Milwaukee, 600 E Greenfield Ave., Milwaukee, WI 53204, USA
a r t i c l e i n f o
a b s t r a c t
Article history: Received 28 February 2018 Received in revised form 10 May 2018 Accepted 31 May 2018 Available online 31 May 2018
Stability of colloidal particles contributes to the turbidity in the water column, which significantly influences water quality and ecological functions in aquatic environments especially shallow lakes. Here we report characterization, origin and aggregation behavior of aquatic colloids, including natural colloidal particles (NCPs) and total inorganic colloidal particles (TICPs), in a highly turbid shallow lake, via field observations, simulation experiments, ultrafiltration, spectral and microscopic, and light scattering techniques. The colloidal particles were characterized with various shapes (spherical, polygonal and elliptical) and aluminum-, silicon-, and ferric-containing mineralogical structures, with a size range of 20e200 nm. The process of sediment re-suspension under environmentally relevant conditions contributed 78e80% of TICPs and 54e55% of NCPs in Lake Taihu, representing an important source of colloids in the water column. Both mono- and divalent electrolytes enhanced colloidal aggregation, while a reverse trend was observed in the presence of natural organic matter (NOM). The influence of NOM on colloidal stability was highly related to molecular weight (MW) properties with the high MW fraction exhibiting higher stability efficiency than the low MW counterparts. However, the MW-dependent aggregation behavior for NCPs was less significant than that for TICPs, implying that previous results on colloidal behavior using model inorganic colloids alone should be reevaluated. Further studies are needed to better understand the mobility/stability and transformation of aquatic colloids and their role in governing the fate and transport of pollutants in natural waters. © 2018 Published by Elsevier Ltd.
Keywords: Eutrophic shallow lakes Aquatic colloids Aggregation behavior Natural organic matter Molecular size fractionation
1. Introduction Natural colloids, defined as particles with sizes ranging between 1 nm and 1 mm, are ubiquitous in aquatic environments (Buffle and Leppard, 1995; Gibson et al., 2007). Due to the small sizes, these colloidal particles usually have a long residence time and substantially impact transparency and recreational value of waters. The properties of high specific surface area and reactivity endow them strong adsorption potential to many trace elements, including nutrients and pollutants, and as a result, influence their chemical speciation, behavior and environmental fate (Gibson et al., 2007). Moreover, the persistence of aquatic colloids can also cause
* Corresponding author. E-mail address: [email protected] (H. Xu). https://doi.org/10.1016/j.watres.2018.05.059 0043-1354/© 2018 Published by Elsevier Ltd.
potential toxicity to microorganism (Shang et al., 2017) and impacts on aquatic ecosystems (Colman et al., 2014. Owing to the significant influence on the fate and transport of pollutants and on ecosystem balance, characterization of aquatic colloids have received increasing attention over the past years. Environmental colloidal particles in lake waters originate from either allochthonous or autochthonous sources, depending on human activity and local meteorological/hydrological conditions. Compared with deep lakes, shallow lakes are usually characterized by higher turbidity and colloidal abundance due to high primary production and sediment re-suspension (Xing and Kong, 2007; Zheng et al., 2015). The size, composition, and spatio-temporal distribution of lake water colloids had previously been reported, but mainly focused on inorganic colloidal particles (ICPs) (Chanudet and Filella, 2007, 2009; Filella et al., 2009). However, ICPs in natural waters are usually coated with dissolved organic
H. Xu et al. / Water Research 142 (2018) 176e186
€fer et al., ligands to form organic-inorganic colloidal particles (Scha 2007). In addition, dissolved organic matter in natural waters had guen et al., been shown to occur mostly in the form of colloids (Gue 2006; Guo et al., 2009; Shirokova et al., 2013; Stolpe et al., 2013). Therefore, the organic, inorganic and organic-inorganic colloids collectively constitute the abundance of natural colloidal particles (NCPs) in lake waters. Nevertheless, knowledge on the properties and sources of colloidal particles (including NCPs and ICPs) in lake waters, especially the high-turbidity shallow lakes, is lacking, but vital for understanding the roles of colloidal particles in regulating water quality, ecosystem function and pollutant transport. The environmental behaviors of colloids in natural waters are controlled by their stability/aggregation propensities, which are highly dependent on environmental conditions. Many parameters, including electrolytes, ionic strength, pH, and natural organic matter (NOM), have been reported to influence the stability of inorganic colloids like kaolinite (Kretzschmar et al., 1998; Aurell and Wistrom, 2000; Sequaris, 2010), montmorillonite (GarciaGarcia et al., 2007; Sequaris, 2010; Borgnino, 2013), illite and quartz (Jiang et al., 2012). Among them, electrolytes and NOM, including quantity and composition, play critical roles in affecting the stability of colloids (Philippe and Schaumann, 2014). It was generally found that electrolytes (including mono- and divalent) enhanced colloidal aggregation, but presence of NOM can significantly inhibit the aggregation potential of colloids/nanoparticles (Chen et al., 2006; Dong and Lo, 2013). However, the NOM-induced aggregation inhibition can be highly related to its inherent chemical structure and molecular size. For instance, bovine serum albumins were found to be more effective than humic and fulvic acids in stabilizing the manganese dioxide colloids due to the difference in molecular sizes of the NOM (Huangfu et al., 2013). In fact, NOM in aquatic environments is a highly heterogeneous mixture with various organic components and a continuous size spectra (Philippe and Schaumann, 2014; Xu and Guo, 2017), and different functional groups and molecular weight (MW) fractions may interact differently with colloidal particles. Within the bulk NOM pool, the high MW (HMW) NOM fraction has been shown to play a central role in regulating the concentration and speciation, and hence the fate, transport and bioavailability of many trace elements in aquatic ecosystems (Guo et al., 2002; Alasonati et al., 2010). In addition, recent studies also highlighted the influence of NOM on colloidal stability and found the MWs and functionalities of NOM an important parameter in controlling aggregation kinetics and stability for nanoparticles, including silver (Yin et al., 2015), gold (Louie et al., 2015), fullerene (Shen et al., 2015), and ZnO (Kteeba et al., 2017). However, these studies usually selected one or two synthesized or engineered nanoparticles as the model materials, while the NCPs are polyfunctional and contain compounds with various compositions and structures. In addition, it is not clear in the aggregation heterogeneities between the NCPs and ICPs in natural waters, leaving a knowledge gap regarding the aggregation behaviors of colloids (including NCPs and ICPs) in environmentally relevant conditions. The objectives of this study were to: (1) characterize the physicochemical properties of aquatic colloids, including NCPs and ICPs, in shallow lakes; (2) define the origins of these colloidal particles; (3) explore the NOM- and electrolyte-related aggregation profiles in ambient conditions and reveal the specific MW fractions in NOM matrix that regulate colloidal stability. Both spectral and microscopic techniques were used to characterize the chemical properties of colloidal particles, and a simulation experiment was carried out to quantify the contribution of sediment re-suspension processes to the colloid abundance in lake waters. In addition, ultrafiltration was used to fractionate the bulk NOM into LMW- and HMW-NOM fractions, whose specific effects on colloidal
177
aggregation under mono- and divalent electrolyte conditions were determined via the dynamic light scattering (DLS) technique. 2. Materials and methods 2.1. Collection of aquatic colloids and sediment core samples Lake Taihu, located between 30 550 4000 e31320 5800 N and 119 520 32’’ 120 360 1000 E, is one of the largest shallow lakes in China (mean water depth: 1.9 m) with high turbidity and low transparency in the water column (Qin et al., 2007). Another feature of Lake Taihu is that it has two distinctive ecological regions: macrophyte- and algae-dominant regions. Surface water samples were collected from the two regions (Fig. S1 in the Supporting Information, SI) with an average wind speed (based on 30 min of continuous measurement) of 2, 4, 6 and 8 m/s, respectively. After collection, the samples were transported to the lab within 4 h and settled statically. The particles with size less than 1 mm (sedimentation velocity < 0.3 cm/h) were collected by siphon based on Stokes' law. The supernatants were then ultrafiltered using a stirred cell ultrafiltration unit (Amicon, 8500) and a 1 kDa membrane disc (Millipore, PLAC07610). The retentate with colloidal sizes between 1 nm and 1 mm was carefully collected to represent the NCPs. An aliquot of the retentate was further treated with 30% hydrogen peroxide to obtain total ICPs (TICPs) until no residual organic matters can be detected in the retentates (Dumat et al., 1997). It was noted that the TICPs herein included pristine inorganic colloids and the inorganic particles that coated with organic matters in natural waters. Both the NCPs and TICPs collected finally were freeze-dried and stored in the desiccator before analysis. In addition to the colloidal samples from the water column, sediment core samples were also collected at the same sampling location using a Ø110 mm L500 mm gravity corer (Rigo Co., Ltd., Japan). At least 25 cm of intact sediment cores (2.4 L) were sampled (Fig. S2a in the SI), kept upright and transported to laboratory within 4 h of sampling. 2.2. Sediment re-suspension experiments A Y-shape sediment generation equipment (patent number: CN200410014329.X) was used to simulate the re-suspension process of lake sediment at 2, 4, 6 and 8 m/s, respectively (Fig. S2b in the SI). The system consisted of a series of individual polyethylene tubes with a length of 180 cm and a diameter of 11 cm. Two electric motor-driven propellers were equipped to simulate wind-induced waves: one was placed 120 cm above the sediment surface to mix the water vertically and the other was placed 15 cm above the sediment surface in the incline tube with an angle of 55 to offer sediment re-suspension force. The rotation of the two propellers was controlled by different motors with a frequency ranging from 0 to 20 Hz to simulate various wind speeds (Liu et al., 2015). Before the re-suspension experiment, the sediment cores were carefully transplanted into lower plexiglass tube of Y-shape apparatus. Simulated lake waters (containing 0.15 mM Kþ, 2 mM Naþ, 0.5 mM Ca2þ, 0.4 mM Mg2þ, and 1.5 mM HCO 3 ) (Yan et al., 2014) were added to each tube to obtain a depth of about 1.9 m, the same as the mean depth of Lake Taihu. For each experiment, surface water samples were collected after 5 h of re-suspension process, and the colloidal particles were isolated using the abovementioned ultrafiltration method. Each re-suspension experiment was conducted in duplicate. The difference in colloidal abundance between lake waters and simulation experiments can be used to quantify the contribution of sediment re-suspension to colloids in the water column.
178
H. Xu et al. / Water Research 142 (2018) 176e186
2.3. Molecular size fractionation of NOM samples Fulvic acid (FA, 99%), as a representative of NOM, was purchased from Jonln Biological Technology Co. (China). The ultrafiltration unit was used to fractionate the bulk FA (<0.45 mm cellulose acetate filtration) into fractions with various molecular sizes. The 1 kDa membrane discs were firstly cleaned with 0.05 N NaOH and 0.02 N HCl, followed by thorough flushing with ultrapure water. The membranes were also cleaned with ultrapure water between chemicals. For ultrafiltration, the bulk water samples (300 ml each) were transferred to the stirred cell loaded with a pre-cleaned 1 kDa membrane (Xu and Guo, 2017). During ultrafiltration, samples were under continuous stirring and a N2 pressure maintaining at 345 kPa (50 psi). The initial bulk and permeate solutions were saved for the measurement of dissolved organic carbon (DOC) and other chemical properties to calculate the size or MW distributions between the <1 kDa and >1 kDa fractions. Stock solutions of 0.05 M CaCl2 and 5 M NaCl were prepared by dissolving analytical-grade CaCl2 and NaCl in ultrapure water and filtered through a 0.22 mm filter. In addition, the NCPs and TICPs collected from algae- and macrophyte-dominant lake waters were re-suspended in the ultrapure water to obtain colloidal suspensions at a mass concentration of ~10 mg/L. 2.4. Hydrodynamic diameters and electrophoretic mobility Measurements of hydrodynamic diameters and electrophoretic mobility were conducted at room temperature and neutral pH using a Malvern Zetasizer (Nano ZS, Malvern, UK). A monochromatic coherent He-Ne laser with a fixed wavelength of 633 nm was used as a light source and the intensity of scattered light was measured by a detector at 173 . Each auto-correlation function was accumulated for 6 s and a total of 10 auto-correlations were obtained for each measurement. For colloidal aggregation in the absence of NOM, the mono- and divalent electrolytes with different concentrations (Naþ: 0e1 M, Ca2þ: 0e4 mM) were added into the colloidal solutions (10 mg/L). The suspensions (pH: 7.0 ± 0.1) were briefly mixed (Vortex Genie 2, Fisher Scientific) for no more than 3 s, then one mL of the suspension was introduced into a disposable polystyrene cuvette (Sarstedt, Germany) to track the timedependent changes in hydrodynamic diameters. For aggregation dynamics in the presence of NOM, a specific concentration (e.g., 2 mg-C/L) of bulk and size-fractionated FA sample was added to the colloidal suspension along with the addition of the electrolytes, with a measurement procedure similar to that in the absence of FA samples. As for the electrophoretic mobility analysis, each sample was measured at least 3 times and data were presented as mean ± standard deviation. 2.5. Aggregation kinetics Time-resolved DLS was used to obtain the aggregation profiles of colloidal particles in the absence or presence of bulk and sizefractionated NOM by measuring the variations in hydrodynamic diameters over time (t). The aggregation rate constant, k, is determined from the relative rate of z-average hydrodynamic diameter changes with time (Chen et al., 2006; Chen and Elimelech, 2006):
k∞
1 dDh ðtÞ N0 dt t/0
(1)
where N0 is the initial particle concentration, and Dh(t) is the hydrodynamic diameter of colloidal particles at time t. The initial rate of increase in Dh(t) is obtained by determining the slope in the
linear regime during the initial 20 min. Since colloidal aggregation was highly dependent on the interaction energy, two aggregation regimes, that is, reaction and diffusion limited aggregation can be identified. Diffusion limited aggregation usually occurs when there is a low or negative energy barrier, under which the repulsive force is negligible and thus the aggregation rate is dependent solely on the time for clusters to encounter each other by diffusion. The reaction limited aggregation, on the other hand, occurs when there is a substantial but not insurmountable energy barrier, under which repulsive and attraction force were highly dependent upon the solution chemistry (e.g., ionic strength, electrolyte valence). Based on different aggregation regimes, the particle attachment efficiency (a) can be obtained to quantitatively compare the aggregation potential of colloidal particles under various conditions. It is defined as normalizing the measured k by those under diffusion limited condition, where the aggregation behavior is independent of the electrolyte concentration. As the concentrations of colloids in all aggregation experiments are the same, the value of a in the absence of NOM is calculated by the initial slope of the aggregation profile at a given electrolyte concentration normalized by kfast under the diffusion limited condition, whereas the a value in the presence of NOM was calculated by normalizing the early stage aggregation rate constants by kfast determined for the respective samples in the absence of NOM (Chen and Elimelech, 2007; Baalousha et al., 2013):
a¼
dDh ðtÞ dt
k ¼ kfast dDh ðtÞ dt
t/0
(2)
t/0; fast
where the kfast is the aggregation rate constant of colloidal particles under the diffusion limited condition. 2.6. Sorption of bulk and MW-fractionated NOM onto aquatic colloidal particles Batch sorption experiments were performed in duplicate to estimate the sorption capacity of bulk and MW-fractionated FA by NCPs and TICPs. Detailed procedures of the sorption experiment have been described elsewhere (Xu and Jiang, 2015) and given in the SI. 2.7. Other chemical analyses DOC concentrations in the bulk and MW-fractionated FA samples were measured by a TOC-VCPH analyzer (TOC-400, Shimadzu). UV absorption spectra were measured using a spectrophotometer (UV-2550, Shimadzu) over a wavelength range of 200e800 nm with 1 nm increments. Fluorescence EEM spectra were measured using a Cary Eclipse fluorescence spectrophotometer (Varian, USA). The spectra were gathered with subsequent scanning emission spectra from 250 to 550 nm at 2 nm increments by varying the excitation wavelength from 200 to 450 nm at 10 nm increments. The spectra were recorded at a scan rate of 1200 nm/min, using excitation and emission slit bandwidths of 5 nm. Observations on HR-TEM were conducted using a JEM 2100 field emission gun TEM (JOEL, Japan). The HR-TEM, selected area electron diffraction and energy dispersive X-ray (EDX) analyses were conducted using a 200 keV electron beam to characterize the sample appearance and structure. Powder XRD patterns were collected on a X-ray diffractometer (Siemens D5000) with Cu-Ka radiation 40 kV and 30 mA with a step size of 0.02 . In addition, the independent-sample T-test
H. Xu et al. / Water Research 142 (2018) 176e186
model was performed to examine the significance of differences between different sample sets. 3. Results and discussion 3.1. Characterization of aquatic colloids Fig. 1 shows typical TEM micrographs and EDX spectra for the colloids collected from different ecological regions of the lake. No significant difference in shapes was observed for NCPs between the two regions. They both contained a mixture of spherical, polygonal and elliptical-shaped colloids with an uneven size distribution, ranging from 20 nm to 200 nm. In addition, some organic-coated boundary can be found on the surface of NCPs, which would be attributed to the chemical and/or physical sorption of macromol€fer et al., 2007). The ecules (Buffle and Leppard, 1995; Scha conformation of TICPs, however, showed an origin-dependent feature. Specifically, TICPs from the macrophyte-dominant region were mostly polygonal, while those from algae-dominant lake waters were mainly spherical. Compared with the wider size range of NCPs, the individual TICPs from the two sampling regions contained mainly colloids with sizes between 20 and 40 nm (Fig. 1), showing a more uniform size distribution for the TICPs. The electron diffraction patterns demonstrated obvious crystalline properties, and chemical composition obtained from EDX analysis indicated that these TICPs contained similar major elements such as Si, Fe, Al, K, and Mg (Table S1 in the SI). Detailed analysis, however, showed that the TICPs from algae-dominant region exhibited higher mineral Al and Si contents than that from macrophytedominant region, which may be responsible for their different conformations. Results from XRD analysis further identified that the main mineral structure for the TICPs included quartz (SiO2), montmorillonite [Na0.3(Al, Mg)2Si4O10(OH)2$8H2O], nontronite [Na0.3Fe2Si4O10(OH)2$4H2O], and volkonskoite [Ca0.3(Cr, Mg)2(Si, Al)4O10(OH)2$4H2O] (Fig. S3 in the SI). These results indicated that Al-, Si-, and Fe-containing nanoparticles are the main mineralogical composition for the TICPs in lake waters. 3.2. Sources of colloidal particles in lake waters Fig. 2 compares the wind-induced concentrations of colloidal particles between natural waters and simulated experiments. The concentrations of TICPs in lake waters ranged from 6.3 ± 0.3 to 20.7 ± 2.2 mg/L and from 2.8 ± 0.2 to 11.4 ± 0.5 mg/L for the algaeand macrophyte-dominant regions, respectively, which were significantly higher than those detected in some deep lake waters (Filella et al., 2009). The concentrations of NCPs were about two times higher than those of TICPs in our samples, indicating the presence and importance of colloidal organic matter in the water column, which is consistent with previous observations for other freshwater systems (Zou et al., 2006; Stolpe et al., 2013; Cai et al., 2015). In addition, adsorption of macromolecules on inorganic colloid surfaces as well as the excretion of colloid-like organic substances from algae and macrophytes could also contribute to the high abundance of NCPs in natural waters (Xu et al., 2013; Xu and Jiang, 2015). Further analysis showed that concentrations of colloidal particles in both lake waters and simulated resuspension experiments increased with increasing wind speeds (Fig. 2a), demonstrating the important role of sediment re-suspension in the formation of colloids in shallow lake waters. In the algae-dominant region, as the wind speed increased from 2 to 8 m/s, the concentrations of NCPs and TICPs in simulated experiments increased from 12.5 ± 0.9 and 4.9 ± 0.3 mg/L to 30.2 ± 2.1 and 16.0 ± 1.1 mg/L, respectively. However, under the same wind speed conditions, the NCPs and TICPs
179
collected from lake waters varied from 23.1 ± 1.9 and 6.3 ± 0.3 mg/L to 58.4 ± 4.2 and 20.7 ± 2.2 mg/L, respectively. The differences in colloidal abundance between lake waters and simulated experiments demonstrated that 53.9 ± 1.8% of NCPs and 78.3 ± 1.1% of TICPs in lake waters could be derived from the sediment resuspension during windy conditions. The effect of sediment resuspension to the aquatic colloid abundance in macrophytedominant region (Fig. 2b) showed similar results that the contribution of sediment re-suspension to natural colloid abundance was 55.0 ± 3.5% for NCPs and 80.1 ± 2.6% for TICPs. The lower percentage contribution of sediment re-suspension for NCPs can be attributed to the presence of colloid-like organic substances in the water column and/or the adsorption of macromolecules on inorganic colloid surfaces.
3.3. Aggregation of aquatic colloids in electrolyte solutions without NOM Time-resolved aggregation profiles of aquatic colloids, including NCPs and TICPs isolated from lake waters, as a function of electrolyte concentrations are presented in Fig. S4 in the SI. Based on aggregation kinetics, the electrolyte-dependent attachment efficiencies can be obtained. As shown in Fig. 3, the aggregation behavior of NCPs and TICPs in both Naþ and Ca2þ solutions exhibited clearly a DLVO-type aggregation profile, suggesting that electrostatic interactions played an important role in stabilizing the aquatic colloids (Yin et al., 2015). This cation-induced aggregation behavior has also been reported for other bare and surfacemodified inorganic colloidal particles (Liu et al., 2011; Dong and Lo, 2013). In this study, the attachment efficiency exhibited an initial electrolyte-dependent enhancement followed by a nonsensitive stage with increasing electrolyte concentrations. The minimum electrolyte concentration required for the occurrence of this non-sensitivity was defined as critical coagulation concentration (CCC) (Chen et al., 2006). It was shown that the CCC values were highly dependent on the valence of electrolytes. Taken the TICPs as an example, the CCCs in Naþ-treatment were 18 ± 2 and 28 ± 2 mM for TICPs from the macrophyte- and algae-dominant regions, respectively, while those in Ca2þ-treatment decreased to 0.55 ± 0.05 and 0.50 ± 0.05 mM, respectively. The much lower CCC value of Ca2þ was ascribed to the high charge density and strong neutralization abilities of divalent electrolytes (Stankus et al., 2011). Since the CCC values of NaCl observed in this study were in the range of 18e28 mM for TICPs, those of CaCl2, calculated based on the Schulze-Hardy rule (Verwey et al., 1999), were thus within 0.28 and 0.44 mM, which were in the same range as the experimental data (Fig. 3). In addition to electrolyte valences, types of colloidal particles also played an important role in determining the CCC values. For example, the NCPs exhibited higher CCC values than TICPs both for the macrophyte- and algae-dominant lake waters. Specifically, the CCC values of Naþ and Ca2þ were 18e28 and 0.50e0.55 mM for TICPs, respectively, which were obviously lower than those for the NCPs (33e150 mM in Naþ and 0.75e1.60 mM in Ca2þ). Based on these CCCs, it is clear that NCPs were more stable or less sensitive to the effect of cations than TICPs in both mono- and divalent electrolyte solutions. The zeta-potentials of TICPs in algae- and macrophyte-dominant lake waters were 42.0 ± 0.3 and 38.5 ± 1.6 mV, respectively, while those of NCPs were in the range from 60.3 ± 1.8 to 60.9 ± 2.3 mV. Lower zeta-potentials and steric repulsion originated from the surface coating suggest that more electrolyte cations are needed to screen the negative charge for colloidal aggregation, which accounts for the higher CCC values of Naþ and Ca2þ for NCPs.
180
H. Xu et al. / Water Research 142 (2018) 176e186
Fig. 1. TEM micrograph of the NCPs (aeb) and TICPs (ced) from macrophyte- and algae-dominant lake waters, respectively. The electron diffraction patterns and EDX maps for the TICPs are shown at the bottom for chemical composition determination.
H. Xu et al. / Water Research 142 (2018) 176e186
181
Fig. 2. The re-suspension process of algae- (a) and macrophyte-dominant (b) lake sediments under different wind speeds and its contribution to the colloidal abundance in overlaying waters.
Fig. 3. Electrolyte-dependent variations in attachment efficiencies of colloidal particles (10 mg/L) collected from macrophyte-(a) and algae-(b) dominant lake waters. The critical coagulation concentration (CCC) values for each situation are marked for comparisons.
3.4. Aggregation kinetics of colloids in electrolyte solution with bulk and MW-fractionated NOM 3.4.1. Characterization of bulk and MW-fractionated NOM Information on MW distribution of FA was obtained via the application of ultrafiltration. Comparisons in DOC concentrations between the bulk and ultrafiltered samples showed that the FA tested was composed of 40% (in mass) of the <1 kDa LMW and 60% of the HMW-FA (1 kDa~0.45 mm). The percentage of the LMW fraction was similar to those observed in natural river NOM samples using the same membrane disc and MWCO (Xu and Guo, 2017) but much lower than that of the Western Siberia lakes (Shirokova et al., 2013). Spectroscopic data for comparing the inherent chemical properties between the bulk and MW-fractionated FA samples are shown in Fig. 4. The absorbance intensities of the >1 kDa HMW-FA were generally higher than those of bulk and LMW-FA at all wavelengths (Fig. 4a). This demonstrated that the aromatic components and hydrophobic structures were partitioned mainly in the HMW fraction, similar with other previous studies (Shen et al.,
2015; Yin et al., 2015). The fluorescent EEM spectra in the bulk FA sample detected four major peaks with Ex/Em maxima at 275/ (320e340), 230/(320e340), 275/(380e400) and 230/(380e400) nm representing peaks A, B, C, D, respectively (Fig. 4b), manifesting the abundance of carboxylic-like functional groups (Yan et al., 2013; Xu et al., 2016a). Among these peaks, peaks A and B with low aromaticity were substaintially located in the LMW-fraction (Fig. 4c), while the HMW-fraction with high aromaticity and hydrophobicity contained mainly peaks C and D (Fig. 4d), showing a MW-dependent feature in abundance and chemical composition in the bulk FA. Similar distribution patterns were also found in other natural waters (Xu and Guo, 2017). 3.4.2. NOM-molecular-size-dependent aggregation behavior of TICPs Based on the time-resolved aggregation profiles (Fig. S5-S8 in the SI), the variations in attachment efficiencies of TICPs as a function of different electrolyte concentrations in the presence of bulk and MW-fractioned FAs can be obtained (Fig. 5). Overall, the addition of FA matrix resulted in the increase in CCC values in both
182
H. Xu et al. / Water Research 142 (2018) 176e186
Fig. 4. Comparison in absorbance (a) and fluorescent (bed) spectra between the bulk and MW-fractionated FA on the same mass basis, 10 mg/L of DOC for each FA fraction.
mono- and divalent electrolytes, showing a FA-induced enhancement in colloidal stability (Dong and Lo, 2013). Specifically, due to the addition of FA matrix, the CCCs of Naþ shifted from 18 ± 2e28 ± 2 mM to 150 ± 5e170 ± 10 mM, and those of Ca2þ increased from 0.50 ± 0.05e0.55 ± 0.05 mM to 3.30 ± 0.05e4.20 ± 0.05 mM. The CCC values in the presence of bulk FA observed herein were comparable with previous studies showing a CCC of 170 mM for Naþ and 2.5 mM for Ca2þ for the aggregation of Al2O3 colloids in the presence of microbially derived extracellular polymeric substance (EPS) (Xu et al., 2016b). The increase in CCC values indicated that colloidal particles were prone to be more stable even with a low DOC concentration, e.g., ~2 mg-C/L. However, in some shallow lakes especially the eutrophic shallow lakes, DOC concentrations can reach up to about 20 mg-C/L (Xu and Jiang, 2013), which were significantly higher than those applied in the simulation study. Due to the negative charge imparted by organic matters, colloids in such high DOC environments are expected to be relatively stable. This mechanism could result in the persistence of high water turbidity in some eutrophic shallow lakes, such as Lakes Taihu, Chaohu, and Dianchi in China (Xing and Kong, 2007; Zhang et al., 2014). In addition to bulk FA, variations in attachment efficiencies of TICPs were also related to the inherent MW properties. For example, CCCs values with the addition of LMW-FA were 2.7e3.4 times higher for Naþ and 2.2e3.0 times higher for Ca2þ, respectively, compared to those without FA. In contrast, the CCCs in the presence of HMW-FA were 7.9e15.6 times higher for Naþ and
12.2e12.4 times higher for Ca2þ, respectively. These results clearly demonstrated that the HMW-FA enhanced colloidal stability much more efficiently than the LMW-FA. Similar with the aggregation profiles of TICPs, the NOM-induced stability enhancement has been reported for some model inorganic colloids or nanoparticles, such as nano zero-valent iron, titanium dioxide, nickel oxide, zinc oxide, silica colloidal particles (Zhang et al., 2009; Dong and Lo, 2013; Kteeba et al., 2017). However, the relationship between specific NOM MW ranges and colloidal stability enhancement has not been well quantified. Our findings here on MW-dependent aggregation heterogeneity in different electrolytes implied that previous observations of NOM-enhanced colloidal stability should mainly resulted from the HMW fraction of NOM matrix. To probe into mechanisms responsible for the stability enhancement, zeta potentials of TICPs in the presence of different electrolytes altered by the bulk and MW-fractionated FA were compared (Fig. S9 in the SI). As depicted, the zeta potential of TICPs was generally more negative in the presence of FA, even with the presence of divalent electrolytes. In addition, the zeta potential of TICPs in the presence of HMW-FA became even more negative compared to those in the presence of bulk or LMW-FA, except for the condition of high Naþ concentrations (Fig. S9 in the SI). The more negative surface charge indicated strong electrostatic repulsion, which accounts for the higher colloidal stability in the presence of HMW-FA. In addition, owing to the plenty of aromatic substances and hydrophobicity, the HMW-FA would be expected to have preferable adsorption on TICPs through hydrophobic and p-p
H. Xu et al. / Water Research 142 (2018) 176e186
183
Fig. 5. The attachment efficiency of TICPs from macrophyte- (a, b) and algae-dominant (c, d) lake waters in the presence of electrolyte solutions as a function of the bulk and MWfractionated FA. The attachment efficiency of TICPs without FA addition and the critical coagulation concentration (CCC) for each situation were also shown for comparisons. The concentrations of TICPs were 10 mg/L, while those of the bulk and MW-fractionated FA were 2 mg/L.
interactions (Yin et al., 2015). Our adsorption experiments also showed that HMW-FA exhibited higher adsorption intensities than the bulk and LMW-FA (Table S2 in the SI). High adsorptive ability of HMW-FA indicated the efficient formation of organic layers on the surface of TICPs, thus enhancing the zeta potential and colloidal stability via steric repulsion (Shen et al., 2015). Therefore, both the electrostatic repulsion and steric hindrance can account for the MW-dependent enhancement of colloidal stability for TICPs in lake waters. 3.4.3. Aggregation kinetics of NCPs affected by bulk and MWfractionated NOM Fig. 6 shows the MW-dependent aggregation behavior of NCPs in the presence of different electrolytes, with the detailed timeresolved aggregation profiles shown in Figs. S5-S8 in the SI. Compared with the electrolyte-induced aggregation profiles in the absence of FA, addition of bulk FA resulted in higher CCC values: 1.2e1.3 times higher for Naþ and 1.3e1.4 times higher for Ca2þ. As for the MW-fractionated FA, the LMW-FA increased the CCC values at 1.2e1.4 times higher for Naþ and 1.1e1.5 times higher for Ca2þ, respectively, and the addition of HMW-FA also caused a slight increase in CCC values at 1.4e1.5 times higher for Naþ and 1.3e1.5 times higher for Ca2þ, respectively. These results demonstrated that the increase extents of CCCs for NCPs were much less significant compared to those for TICPs both in mono- and divalent electrolytes due to the difference in colloidal composition. Yin et al. (2015) also showed that NOM enhanced the stability of bare silver nanoparticles but exhibited minor influence for the citrate-coated nanoparticles (~20 nm). However, another study reported that addition of NOM elevated the stability of both citrate- and polyvinylpyrrolidone-coated silver nanoparticles (~60 nm) in
solutions containing Naþ or low concentrations of Ca2þ, but inhibited the stability at high Ca2þ concentrations (Huynh and Chen, 2011). This indicated that the influences of NOM on the aggregation profiles of surface-coated colloidal particles also depended on other factors, such as colloid properties, coating agents, and specific electrolytes. Our results were consistent with those of Yin et al. (2015). To reveal the mechanism for the insensitivity of aggregation kinetics in response to FA addition, the surface charges and Fourier transform infrared (FTIR) spectra of NCPs under various conditions were measured. The zeta potential and FTIR spectra of NCPs in both mono- and divalent electrolytes was not significantly influenced by the bulk and MW-fractionated FA (Figs. S10 and S11 in the SI), demonstrating weak interactions between FA and the coated organic layers on NCP surfaces. The adsorption experiments also showed that the adsorption capacities of bulk and MW-fractionated FA on NCPs surfaces were generally lower (p < 0.05) than those on TICPs (Table S2 in the SI). Therefore, the presence of organic layers inhibited the further adsorption (or coating) of the bulk and MWfractionated FA on the surfaces of NCPs, which were responsible for the slight MW-dependent enhancement of CCC values. 3.5. Significance and implications Due to the high surface activity and potential influence on the toxicity and bio-availability of many contaminants, aquatic colloids have received increasing attention in recent years (Lead and Wilkinson, 2006; Chanudet and Filella, 2009; Filella et al., 2009; Louie et al., 2016). However, many previous studies focused mainly on the spatio-temporal distribution and abundance analysis, lacking the information on sources of the colloidal particles (Pokrovsky
184
H. Xu et al. / Water Research 142 (2018) 176e186
Fig. 6. Variations in attachment efficiencies of aquatic NCPs from macrophyte- (a, b) and algae-dominant (c, d) lake waters in the presence of bulk and MW-fractionated FA. The attachment efficiency of NCPs without FA and the critical coagulation concentration (CCC) for each situation were also provided for comparisons. The concentrations of NCPs were 10 mg/L, while those of the bulk and MW-fractionated FA were 2 mg/L.
et al., 2016). In this study, a sediment generation apparatus was applied to quantify the contribution of sediment re-suspension to the colloid abundance in natural waters. It was found that, under environmentally relevant conditions, the process of sediment resuspension contributed up to 78e80% of TICPs and 54e55% of NCPs in the water column of eutrophic shallow lakes. To our knowledge, this is the first study elucidating the sources and production of natural colloids. As a ubiquitous entity, NOM can regulate the stability of colloids via electrostatic repulsion and steric hindrance (Furman et al., 2013; Philippe and Schaumann, 2014). Although previous studies on the NOM effects on colloidal stability were mainly based on model inorganic colloids, a consistent NOM-induced stability enhancement has been observed in both mono- and divalent electrolytes, similar to those observed for TICPs here. However, we found that the NOM-induced aggregation heterogeneities of the NCPs were less significant than those of the TICPs. As known, engineered nanoparticles are likely to be coated with organic macromolecules, either intentionally added by manufacturers for desired properties or incidentally obtained during their life cycle (Louie et al., 2013). Moreover, the large specific surface area endowed inorganic colloids a high capacity for NOM adsorption or coating (Table S2 in the SI), forming the predominant NCPs species in aquatic environments. Thus, due to the noticeable heterogeneities in aggregation profiles between the NCPs and TICPs, previous results on colloidal behavior analysis using model inorganic colloids are likely to deviate from those of aquatic colloids and may need a critical reevaluation. Furthermore, since the bulk NOM is highly heterogeneous in molecular size and composition, different molecular size fractions
of NOM are expected to play distinct roles in determining the aggregation behavior. Compared with previous studies aiming at the effects of bulk NOM (Zhang et al., 2009; Dong and Lo, 2013), we found evident MW-dependent aggregation heterogeneity in both mono- and divalent electrolytes with the >1 kDa HMW-FA having higher stability efficiency, followed by bulk FA and then the <1 kDa LMW-FA. This is consistent with the distribution of aromatic and hydrophobic organic components among the LMW-, HMW- and bulk NOM pools (Cai et al., 2015). The results highlighted the heterogeneity of NOM in regulating colloidal stability among different MW fractions within NOM matrix. Overall, our results revealed that even a low concentration of organic ligands (~2 mg-C/L) can significantly enhance the colloidal stability, and divalent cations enhanced colloidal stability more efficiently than monovalent cations. As reported, fresh waters usually contain a Ca2þ concentration less than 1.0 mM and Naþ less than 2.5 mM (Xu et al., 2016b). This indicated that the colloidal particles, including NCPs and TICPs, in most freshwater systems, were usually characterized as obvious reaction-limited aggregation regime. This also accounted for the persistence of high water turbidity in many aquatic ecosystems. The mechanisms for colloidal aggregation in natural waters included electrostatic neutralization and/or interparticle bridging (Huangfu et al., 2013; Philippe and Schaumann, 2014). Our previous studies reported that, in the presence of cyanobacterial EPS, electrostatic neutralization contributed to the aggregation of model inorganic colloids at low Ca2þ concentration, while cation bridging became the main mechanism for the enhanced aggregation in Ca2þ concentrations higher than the CCC (Xu et al., 2016b). Yin et al. (2015) also found that although HMW-NOM was the key
H. Xu et al. / Water Research 142 (2018) 176e186
component in stabilizing the citrate-coated silver colloids at low Ca2þ concentrations, it could play a reversed role in promoting the colloidal aggregation at high concentrations of divalent cations. However, the phenomenon of the interparticle bridging was not observed in this study both for the NCPs and TICPs. The reason for the absence of bridging mechanism was still unclear, but may be related to factors, such as pH, NOM composition, molecular size, and colloidal structure, which needs further studies. 4. Conclusion Aquatic colloids, including NCPs and TICPs, in eutrophic shallow lake waters, were characterized with high abundance, broad size distribution, and Fe/Al/Si-containing mineralogical properties. Sediment re-suspension can have a significant contribution to colloidal particles at 54e55% of the NCPs and 78e80% of the TICPs in lake waters. The colloidal particles exhibited obvious DLVO-like aggregation profiles in mono- and divalent electrolytes, but presence of FA significantly inhibited the aggregation and enhanced the colloidal stability in the water column. Different MW fractions within FA matrix showed different effects on colloidal stability against aggregation with HMW fraction exhibiting higher stability efficiency than the LMW counterpart. However, the MWdependent aggregation heterogeneities for NCPs were less significant than those for TICPs. Therefore, in addition to bulk NOM, influence of DOM with different size fractions within NOM matrix should also be considered when interpreting effects of NOM on colloidal behavior. In addition differences in the NOM MWdependent aggregation behavior between NCPs and TICPs indicated that previous results on colloidal behavior analysis using model inorganic colloids should be reevaluated. Acknowledgements We gratefully thank three anonymous reviewers for their constructive comments and suggestions which improved the manuscript. This study was supported, in part, by the National Natural Science Foundation of China (51479187) and Youth Innovation Promotion Association CAS (2016286). We also thank Dr. Cheng Liu and Qiushi Shen for their help in sediment re-suspension simulation. Appendix A. Supplementary data Supplementary data related to this article can be found at https://doi.org/10.1016/j.watres.2018.05.059. References Alasonati, E., Slaveykova, V.I., Gallard, H., Croue, J.P., Benedetti, M.F., 2010. Characterization of the colloidal organic matter from the Amazonian basin by asymmetrical flow field-flow fractionation and size exclusion chromatography. Water Res. 44 (1), 223e231. Aurell, C.A., Wistrom, A.O., 2000. Coagulation of kaolinite colloids in high carbonate strength water. Colloid. Surface. Physicochem. Eng. Aspect. 168 (3), 277e285. Baalousha, M., Nur, Y., Romer, I., Tejamaya, M., Lead, J.R., 2013. Effect of monovalent and divalent cations, anions and fulvic acid on aggregation of citrate-coated silver nanoparticles. Sci. Total Environ. 454, 119e131. Borgnino, L., 2013. Experimental determination of the colloidal stability of Fe(III)montmorillonite: effects of organic matter, ionic strength and pH conditions. Colloid. Surface. Physicochem. Eng. Aspect. 423, 178e187. Buffle, J., Leppard, G.G., 1995. Characterization of aquatic colloids and macromolecules. 1. structure and behavior of colloidal material. Environ. Sci. Technol. 29 (9), 2169e2175. Cai, Y.H., Guo, L.D., Wang, X.R., Aiken, G., 2015. Abundance, stable isotopic composition, and export fluxes of DOC, POC, and DIC from the Lower Mississippi River during 2006e2008. J. Geophys. Res.: Biogeosci. 120 (11), 2273e2288. Chanudet, V., Filella, M., 2007. The fate of inorganic colloidal particles in Lake Brienz. Aquatic Sci.-Res. Across Boundaries 69 (2), 199e211.
185
Chanudet, V., Filella, M., 2009. Size and composition of inorganic colloids in a perialpine, glacial flour-rich lake. Geochem. Cosmochim. Acta 72 (5), 1466e1479. Chen, K.L., Elimelech, M., 2006. Aggregation and deposition kinetics of fullerene (C60) nanoparticles. Langmuir 22 (26), 10994e11001, 2006. Chen, K.L., Mylon, S.E., Elimelech, M., 2006. Aggregation kinetics of alginate-coated hematite nanoparticles in monovalent and divalent electrolytes. Environ. Sci. Technol. 40 (5), 1516e1523. Chen, K.L., Elimelech, M., 2007. Influence of humic acid on the aggregation kinetics of fullerene (C60) nanoparticles in monovalent and divalent electrolyte solutions. J. Colloid Interface Sci. 309 (1), 126e134. Colman, B.P., Espinasse, B., Richardson, C.J., Matson, C.W., Lowry, G.V., Hunt, D.E., Wiesner, M.R., Bernhardt, E.S., 2014. Emerging contaminant or an old toxin in disguise? Silver nanoparticle impacts on ecosystems. Environ. Sci. Technol. 48 (9), 5229e5236. Dong, H.R., Lo, I.M.C., 2013. Influence of humic acid on the colloidal stability of surface-modified nano zero-valent iron. Water Res. 47 (1), 419e427. Dumat, C., Cheshire, M.V., Fraser, A.R., Shand, C.A., Staunton, S., 1997. The effect of removal of soil organic matter and iron on the adsorption of radiocaesium. Eur. J. Soil Sci. 48 (4), 675e683. Filella, M., Chanudet, V., Philippo, S., Quentel, F., 2009. Particle size and mineralogical composition of inorganic colloids in waters draining the adit of an abandoned mine, Goesdorf, Luxembourg. Appl. Geochem. 24 (1), 52e61. Furman, O., Usenko, S., Lau, B.L.T., 2013. Relative importance of the humic and fulvic fractions of natural organic matter in the aggregation and deposition of silver nanoparticles. Environ. Sci. Technol. 47 (3), 1349e1356. Garcia-Garcia, S., Wold, S., Jonsson, M., 2007. Kinetic determination of critical coagulation concentrations for sodium- and calcium-montmorillonite colloids in NaCl and CaCl2 aqueous solutions. J. Colloid Interface Sci. 315 (2), 512e519. Gibson, C.T., Turner, I.J., Roberts, C.J., Lead, J.R., 2007. Quantifying the dimensions of nanoscale organic surface layers in natural waters. Environ. Sci. Technol. 41 (4), 1339e1344. guen, C., Guo, L.D., Wang, D.L., Tanaka, N., Hung, C.C., 2006. Chemical characGue teristics and origin of dissolved organic matter in the Yukon River. Biogeochemistry 77 (2), 139e155. Guo, L.D., Santschi, P.H., Ray, S.M., 2002. Metal partitioning between colloidal and dissolved phases and its relation with bioavailability to American oysters. Mar. Environ. Res. 54 (1), 49e64. Guo, L.D., White, D.M., Xu, C., Santschi, P.H., 2009. Chemical and isotopic composition of HMW-DOM from the Mississippi River plume. Mar. Chem. 114, 63e71. Huangfu, X.L., Jiang, J., Ma, J., Liu, Y.Z., Yang, J., 2013. Aggregation kinetics of manganese dioxide colloids in aqueous solution: influence of humic substances and biomacromolecules. Environ. Sci. Technol. 47 (18), 10285e10292. Huynh, K.A., Chen, K.L., 2011. Aggregation kinetics of citrate and polyvinylpyrrolidone coated silver nanoparticles in monovalent and divalent electrolyte solutions. Environ. Sci. Technol. 45 (13), 5564e5571. Jiang, C.L., Sequaris, J.M., Vereecken, H., Klumpp, E., 2012. Effects of inorganic and organic anions on the stability of illite and quartz soil colloids in Na-, Ca- and mixed Na-Ca systems. Colloid. Surface. Physicochem. Eng. Aspect. 415, 134e141. Kretzschmar, R., Holthoff, H., Sticher, H., 1998. Influence of pH and humic acid on coagulation kinetics of kaolinite: a dynamic light scattering study. J. Colloid Interface Sci. 202 (1), 95e103. Kteeba, S.M., El-Adawi, H.I., El-Rayis, O.A., El-Ghobashy, A.E., Schuld, J.L., Svoboda, K.R., Guo, L.D., 2017. Zinc oxide nanoparticle toxicity in embryonic zebrafish: mitigation with different natural organic matter. Environ. Pollut. 230, 1125e1140. Lead, J.R., Wilkinson, K.J., 2006. Aquatic colloids and nanoparticles: current knowledge and future trends. Environ. Chem. 3 (3), 159e171. Liu, C., Shen, Q.S., Zhou, Q.L., Fan, C.X., Shao, S.G., 2015. Precontrol of algae-induced black blooms through sediment dredging at appropriate depth in a typical eutrophic shallow lake. Ecol. Eng. 77, 139e145. Liu, X.Y., Wazne, M., Chou, T.M., Xiao, R., Xu, S.Y., 2011. Influence of Ca2þ and Suwannee river humic acid on aggregation of silicon nanoparticles in aqueous media. Water Res. 45 (1), 105e112. Louie, S.M., Tilton, R.D., Lowry, G.V., 2013. Effects of molecular weight distribution and chemical properties of natural organic matter on gold nanoparticle aggregation. Environ. Sci. Technol. 47 (9), 4245e4254. Louie, S.M., Spielman-Sun, E.R., Small, M.J., Tilton, R.D., Lowry, G.V., 2015. Correlation of the physicochemical properties of natural organic matter samples from different sources to their effects on gold nanoparticle aggregation in monovalent electrolyte. Environ. Sci. Technol. 49 (4), 2188e2198. Louie, S.M., Tilton, R.D., Lowry, G.V., 2016. Critical review: impacts of macromolecular coatings on critical physicochemical processes controlling environmental fate of nanomaterials. Environmental Science: Nano 3 (2), 283e310. Philippe, A., Schaumann, G.E., 2014. Interactions of dissolved organic matter with natural and engineered inorganic colloids: a review. Environ. Sci. Technol. 48 (16), 8946e8962. Pokrovsky, O.S., Manasypov, R.M., Loiko, S.V., Shirokova, L.S., 2016. Organic and organo-mineral colloids in discontinuous permafrost zone. Geochem. Cosmochim. Acta 188, 1e20. Qin, B.Q., Xu, P.Z., Wu, Q.L., Luo, L.C., Zhang, Y.L., 2007. Environmental issues of lake Taihu, China. Hydrobiologia 581 (1), 3e14. €fer, T., Chanudet, V., Claret, F., Filella, M., 2007. Spectromicroscopy mapping of Scha colloidal/particulate organic matter in Lake Brienz, Switzerland. Environ. Sci. Technol. 41 (22), 7864e7869. Sequaris, J.M., 2010. Modeling the effects of Ca2þ and clay-associated organic carbon
186
H. Xu et al. / Water Research 142 (2018) 176e186
on the stability of colloids from topsoils. J. Colloid Interface Sci. 343 (2), 408e414. Shang, E.X., Li, Y., Niu, J.F., Zhou, Y.J., Wang, T.Y., Crittenden, J.C., 2017. Relative importance of humic and fulvic acid on ROS generation, dissolution, and toxicity of sulfide nanoparticles. Water Res. 124, 595e604. Shen, M.H., Yin, Y.G., Booth, A., Liu, J.F., 2015. Effects of molecular weight-dependent physicochemical heterogeneity of natural organic matter on the aggregation of fullerene nanoparticles in mono-and di-valent electrolyte solutions. Water Res. 71, 11e20. Shirokova, L.S., Pokrovsky, O.S., Kirpotin, S.N., Desmukh, C., Pokrovsky, B.G., Audry, S., Viers, J., 2013. Biogeochemistry of organic carbon, CO2, CH4, and trace elements in thermokarst water bodies in discontinuous permafrost zones of Western Siberia. Biogeochemistry 113, 573e593. Stankus, D.P., Lohse, S.E., Hutchison, J.E., Nason, J.A., 2011. Interactions between natural organic matter and gold nanoparticles stabilized with different organic capping agents. Environ. Sci. Technol. 45 (8), 3238e3244. Stolpe, B., Guo, L.D., Shiller, A.M., Aiken, G.R., 2013. Abundance, size distributions and trace-element binding of organic and iron-rich nanocolloids in Alaskan rivers, as revealed by field-flow fractionation and ICP-MS. Geochem. Cosmochim. Acta 105, 221e239. Verwey, E.J.W., Overbeek, J.T.G., Overbeek, J.T.G., 1999. Theory of the Stability of Lyophobic Colloids. Courier Corporation. Xing, P., Kong, F.X., 2007. Intra-habitat heterogeneity of environmental factors regulating bacterioplankton community composition in Lake Taihu, China. Aquat. Microb. Ecol. 48 (2), 113e122. Xu, H.C., Jiang, H.L., 2013. UV-induced photochemical heterogeneity of dissolved and attached organic matter associated with cyanobacterial blooms in a eutrophic freshwater lake. Water Res. 47 (17), 6506e6515. Xu, H.C., Cai, H.Y., Yu, G.H., Jiang, H.L., 2013. Insights into extracellular polymeric substances of cyanobacterium Microcystis aeruginosa using fractionation procedure and parallel factor analysis. Water Res. 47 (6), 2005e2014. Xu, H.C., Jiang, H.L., 2015. Effects of cyanobacterial extracellular polymeric substances on the stability of ZnO nanoparticles in eutrophic shallow lakes.
Environ. Pollut. 197, 231e239. Xu, H.C., Lv, H., Liu, X., Wang, P.F., Jiang, H.L., 2016a. Electrolyte cations binding with extracellular polymeric substances enhanced Microcystis aggregation: implication for Microcystis bloom formation in eutrophic freshwater lakes. Environ. Sci. Technol. 50 (17), 9034e9043. Xu, H.C., Yang, C.M., Jiang, H.L., 2016b. Aggregation kinetics of inorganic colloids in eutrophic shallow lakes: influence of cyanobacterial extracellular polymeric substances and electrolyte cations. Water Res. 106, 344e351. Xu, H.C., Guo, L.D., 2017. Molecular size-dependent abundance and composition of dissolved organic matter in river, lake and sea waters. Water Res. 117, 115e126. Yan, M.Q., Fu, Q.W., Li, D.C., Gao, G.F., Wang, D.S., 2013. Study of the pH influence on the optical properties of dissolved organic matter using fluorescence excitationeemission matrix and parallel factor analysis. J. Lumin. 142, 103e109. Yan, Z.S., Jiang, H.L., Li, X.H., Shi, Y., 2014. Accelerated removal of pyrene and benzo [a]pyrene in freshwatersediments with amendment of cyanobacteria-derived organic matter. J. Hazard Mater. 272, 66e74. Yin, Y.G., Shen, M.H., Tan, Z.Q., Yu, S.J., Liu, J.F., Jiang, G.B., 2015. Particle coatingdependent interaction of molecular weight fractionated natural organic matter: impacts on the aggregation of silver nanoparticles. Environ. Sci. Technol. 49 (11), 6581e6589. Zhang, Y., Chen, Y.S., Westerhoff, P., Crittenden, J., 2009. Impact of natural organic matter and divalent cations on the stability of aqueous nanoparticles. Water Res. 43 (17), 4249e4257. Zhang, Y.L., Shi, K., Liu, X.H., Zhou, Y.Q., Qin, B.Q., 2014. Lake topography and wind waves determining seasonal-spatial dynamics of total suspended matter in turbid Lake Taihu, China: assessment using long-term high-resolution MERIS data. PLoS One 9 (5), e98055. Zheng, S.S., Wang, P.F., Wang, C., Hou, J., 2015. Sediment resuspension under action of wind in Taihu Lake, China. Int. J. Sediment Res. 30 (1), 48e62. Zou, L., Sun, M.Y., Guo, L.D., 2006. Temporal variations of organic carbon inputs into the upper Yukon River: evidence from fatty acids and their stable carbon isotopic compositions in dissolved, colloidal and particulate phases. Org. Geochem. 37 (8), 944e956.
Contents lists available at ScienceDirect
Water Research journal homepage: www.elsevier.com/locate/watres
Characterization, origin and aggregation behavior of colloids in eutrophic shallow lake Huacheng Xu a, *, Mengwen Xu b, Yani Li a, c, Xin Liu b, Laodong Guo d, Helong Jiang a a
State Key Laboratory of Lake Science and Environment, Nanjing Institute of Geography and Limnology, Chinese Academy of Sciences, Nanjing 210008, China b College of Biology and the Environment, Nanjing Forestry University, Nanjing 210037, China c University of Chinese Academy of Sciences, Beijing 100049, China d School of Freshwater Sciences, University of Wisconsin-Milwaukee, 600 E Greenfield Ave., Milwaukee, WI 53204, USA
a r t i c l e i n f o
a b s t r a c t
Article history: Received 28 February 2018 Received in revised form 10 May 2018 Accepted 31 May 2018 Available online 31 May 2018
Stability of colloidal particles contributes to the turbidity in the water column, which significantly influences water quality and ecological functions in aquatic environments especially shallow lakes. Here we report characterization, origin and aggregation behavior of aquatic colloids, including natural colloidal particles (NCPs) and total inorganic colloidal particles (TICPs), in a highly turbid shallow lake, via field observations, simulation experiments, ultrafiltration, spectral and microscopic, and light scattering techniques. The colloidal particles were characterized with various shapes (spherical, polygonal and elliptical) and aluminum-, silicon-, and ferric-containing mineralogical structures, with a size range of 20e200 nm. The process of sediment re-suspension under environmentally relevant conditions contributed 78e80% of TICPs and 54e55% of NCPs in Lake Taihu, representing an important source of colloids in the water column. Both mono- and divalent electrolytes enhanced colloidal aggregation, while a reverse trend was observed in the presence of natural organic matter (NOM). The influence of NOM on colloidal stability was highly related to molecular weight (MW) properties with the high MW fraction exhibiting higher stability efficiency than the low MW counterparts. However, the MW-dependent aggregation behavior for NCPs was less significant than that for TICPs, implying that previous results on colloidal behavior using model inorganic colloids alone should be reevaluated. Further studies are needed to better understand the mobility/stability and transformation of aquatic colloids and their role in governing the fate and transport of pollutants in natural waters. © 2018 Published by Elsevier Ltd.
Keywords: Eutrophic shallow lakes Aquatic colloids Aggregation behavior Natural organic matter Molecular size fractionation
1. Introduction Natural colloids, defined as particles with sizes ranging between 1 nm and 1 mm, are ubiquitous in aquatic environments (Buffle and Leppard, 1995; Gibson et al., 2007). Due to the small sizes, these colloidal particles usually have a long residence time and substantially impact transparency and recreational value of waters. The properties of high specific surface area and reactivity endow them strong adsorption potential to many trace elements, including nutrients and pollutants, and as a result, influence their chemical speciation, behavior and environmental fate (Gibson et al., 2007). Moreover, the persistence of aquatic colloids can also cause
* Corresponding author. E-mail address: [email protected] (H. Xu). https://doi.org/10.1016/j.watres.2018.05.059 0043-1354/© 2018 Published by Elsevier Ltd.
potential toxicity to microorganism (Shang et al., 2017) and impacts on aquatic ecosystems (Colman et al., 2014. Owing to the significant influence on the fate and transport of pollutants and on ecosystem balance, characterization of aquatic colloids have received increasing attention over the past years. Environmental colloidal particles in lake waters originate from either allochthonous or autochthonous sources, depending on human activity and local meteorological/hydrological conditions. Compared with deep lakes, shallow lakes are usually characterized by higher turbidity and colloidal abundance due to high primary production and sediment re-suspension (Xing and Kong, 2007; Zheng et al., 2015). The size, composition, and spatio-temporal distribution of lake water colloids had previously been reported, but mainly focused on inorganic colloidal particles (ICPs) (Chanudet and Filella, 2007, 2009; Filella et al., 2009). However, ICPs in natural waters are usually coated with dissolved organic
H. Xu et al. / Water Research 142 (2018) 176e186
€fer et al., ligands to form organic-inorganic colloidal particles (Scha 2007). In addition, dissolved organic matter in natural waters had guen et al., been shown to occur mostly in the form of colloids (Gue 2006; Guo et al., 2009; Shirokova et al., 2013; Stolpe et al., 2013). Therefore, the organic, inorganic and organic-inorganic colloids collectively constitute the abundance of natural colloidal particles (NCPs) in lake waters. Nevertheless, knowledge on the properties and sources of colloidal particles (including NCPs and ICPs) in lake waters, especially the high-turbidity shallow lakes, is lacking, but vital for understanding the roles of colloidal particles in regulating water quality, ecosystem function and pollutant transport. The environmental behaviors of colloids in natural waters are controlled by their stability/aggregation propensities, which are highly dependent on environmental conditions. Many parameters, including electrolytes, ionic strength, pH, and natural organic matter (NOM), have been reported to influence the stability of inorganic colloids like kaolinite (Kretzschmar et al., 1998; Aurell and Wistrom, 2000; Sequaris, 2010), montmorillonite (GarciaGarcia et al., 2007; Sequaris, 2010; Borgnino, 2013), illite and quartz (Jiang et al., 2012). Among them, electrolytes and NOM, including quantity and composition, play critical roles in affecting the stability of colloids (Philippe and Schaumann, 2014). It was generally found that electrolytes (including mono- and divalent) enhanced colloidal aggregation, but presence of NOM can significantly inhibit the aggregation potential of colloids/nanoparticles (Chen et al., 2006; Dong and Lo, 2013). However, the NOM-induced aggregation inhibition can be highly related to its inherent chemical structure and molecular size. For instance, bovine serum albumins were found to be more effective than humic and fulvic acids in stabilizing the manganese dioxide colloids due to the difference in molecular sizes of the NOM (Huangfu et al., 2013). In fact, NOM in aquatic environments is a highly heterogeneous mixture with various organic components and a continuous size spectra (Philippe and Schaumann, 2014; Xu and Guo, 2017), and different functional groups and molecular weight (MW) fractions may interact differently with colloidal particles. Within the bulk NOM pool, the high MW (HMW) NOM fraction has been shown to play a central role in regulating the concentration and speciation, and hence the fate, transport and bioavailability of many trace elements in aquatic ecosystems (Guo et al., 2002; Alasonati et al., 2010). In addition, recent studies also highlighted the influence of NOM on colloidal stability and found the MWs and functionalities of NOM an important parameter in controlling aggregation kinetics and stability for nanoparticles, including silver (Yin et al., 2015), gold (Louie et al., 2015), fullerene (Shen et al., 2015), and ZnO (Kteeba et al., 2017). However, these studies usually selected one or two synthesized or engineered nanoparticles as the model materials, while the NCPs are polyfunctional and contain compounds with various compositions and structures. In addition, it is not clear in the aggregation heterogeneities between the NCPs and ICPs in natural waters, leaving a knowledge gap regarding the aggregation behaviors of colloids (including NCPs and ICPs) in environmentally relevant conditions. The objectives of this study were to: (1) characterize the physicochemical properties of aquatic colloids, including NCPs and ICPs, in shallow lakes; (2) define the origins of these colloidal particles; (3) explore the NOM- and electrolyte-related aggregation profiles in ambient conditions and reveal the specific MW fractions in NOM matrix that regulate colloidal stability. Both spectral and microscopic techniques were used to characterize the chemical properties of colloidal particles, and a simulation experiment was carried out to quantify the contribution of sediment re-suspension processes to the colloid abundance in lake waters. In addition, ultrafiltration was used to fractionate the bulk NOM into LMW- and HMW-NOM fractions, whose specific effects on colloidal
177
aggregation under mono- and divalent electrolyte conditions were determined via the dynamic light scattering (DLS) technique. 2. Materials and methods 2.1. Collection of aquatic colloids and sediment core samples Lake Taihu, located between 30 550 4000 e31320 5800 N and 119 520 32’’ 120 360 1000 E, is one of the largest shallow lakes in China (mean water depth: 1.9 m) with high turbidity and low transparency in the water column (Qin et al., 2007). Another feature of Lake Taihu is that it has two distinctive ecological regions: macrophyte- and algae-dominant regions. Surface water samples were collected from the two regions (Fig. S1 in the Supporting Information, SI) with an average wind speed (based on 30 min of continuous measurement) of 2, 4, 6 and 8 m/s, respectively. After collection, the samples were transported to the lab within 4 h and settled statically. The particles with size less than 1 mm (sedimentation velocity < 0.3 cm/h) were collected by siphon based on Stokes' law. The supernatants were then ultrafiltered using a stirred cell ultrafiltration unit (Amicon, 8500) and a 1 kDa membrane disc (Millipore, PLAC07610). The retentate with colloidal sizes between 1 nm and 1 mm was carefully collected to represent the NCPs. An aliquot of the retentate was further treated with 30% hydrogen peroxide to obtain total ICPs (TICPs) until no residual organic matters can be detected in the retentates (Dumat et al., 1997). It was noted that the TICPs herein included pristine inorganic colloids and the inorganic particles that coated with organic matters in natural waters. Both the NCPs and TICPs collected finally were freeze-dried and stored in the desiccator before analysis. In addition to the colloidal samples from the water column, sediment core samples were also collected at the same sampling location using a Ø110 mm L500 mm gravity corer (Rigo Co., Ltd., Japan). At least 25 cm of intact sediment cores (2.4 L) were sampled (Fig. S2a in the SI), kept upright and transported to laboratory within 4 h of sampling. 2.2. Sediment re-suspension experiments A Y-shape sediment generation equipment (patent number: CN200410014329.X) was used to simulate the re-suspension process of lake sediment at 2, 4, 6 and 8 m/s, respectively (Fig. S2b in the SI). The system consisted of a series of individual polyethylene tubes with a length of 180 cm and a diameter of 11 cm. Two electric motor-driven propellers were equipped to simulate wind-induced waves: one was placed 120 cm above the sediment surface to mix the water vertically and the other was placed 15 cm above the sediment surface in the incline tube with an angle of 55 to offer sediment re-suspension force. The rotation of the two propellers was controlled by different motors with a frequency ranging from 0 to 20 Hz to simulate various wind speeds (Liu et al., 2015). Before the re-suspension experiment, the sediment cores were carefully transplanted into lower plexiglass tube of Y-shape apparatus. Simulated lake waters (containing 0.15 mM Kþ, 2 mM Naþ, 0.5 mM Ca2þ, 0.4 mM Mg2þ, and 1.5 mM HCO 3 ) (Yan et al., 2014) were added to each tube to obtain a depth of about 1.9 m, the same as the mean depth of Lake Taihu. For each experiment, surface water samples were collected after 5 h of re-suspension process, and the colloidal particles were isolated using the abovementioned ultrafiltration method. Each re-suspension experiment was conducted in duplicate. The difference in colloidal abundance between lake waters and simulation experiments can be used to quantify the contribution of sediment re-suspension to colloids in the water column.
178
H. Xu et al. / Water Research 142 (2018) 176e186
2.3. Molecular size fractionation of NOM samples Fulvic acid (FA, 99%), as a representative of NOM, was purchased from Jonln Biological Technology Co. (China). The ultrafiltration unit was used to fractionate the bulk FA (<0.45 mm cellulose acetate filtration) into fractions with various molecular sizes. The 1 kDa membrane discs were firstly cleaned with 0.05 N NaOH and 0.02 N HCl, followed by thorough flushing with ultrapure water. The membranes were also cleaned with ultrapure water between chemicals. For ultrafiltration, the bulk water samples (300 ml each) were transferred to the stirred cell loaded with a pre-cleaned 1 kDa membrane (Xu and Guo, 2017). During ultrafiltration, samples were under continuous stirring and a N2 pressure maintaining at 345 kPa (50 psi). The initial bulk and permeate solutions were saved for the measurement of dissolved organic carbon (DOC) and other chemical properties to calculate the size or MW distributions between the <1 kDa and >1 kDa fractions. Stock solutions of 0.05 M CaCl2 and 5 M NaCl were prepared by dissolving analytical-grade CaCl2 and NaCl in ultrapure water and filtered through a 0.22 mm filter. In addition, the NCPs and TICPs collected from algae- and macrophyte-dominant lake waters were re-suspended in the ultrapure water to obtain colloidal suspensions at a mass concentration of ~10 mg/L. 2.4. Hydrodynamic diameters and electrophoretic mobility Measurements of hydrodynamic diameters and electrophoretic mobility were conducted at room temperature and neutral pH using a Malvern Zetasizer (Nano ZS, Malvern, UK). A monochromatic coherent He-Ne laser with a fixed wavelength of 633 nm was used as a light source and the intensity of scattered light was measured by a detector at 173 . Each auto-correlation function was accumulated for 6 s and a total of 10 auto-correlations were obtained for each measurement. For colloidal aggregation in the absence of NOM, the mono- and divalent electrolytes with different concentrations (Naþ: 0e1 M, Ca2þ: 0e4 mM) were added into the colloidal solutions (10 mg/L). The suspensions (pH: 7.0 ± 0.1) were briefly mixed (Vortex Genie 2, Fisher Scientific) for no more than 3 s, then one mL of the suspension was introduced into a disposable polystyrene cuvette (Sarstedt, Germany) to track the timedependent changes in hydrodynamic diameters. For aggregation dynamics in the presence of NOM, a specific concentration (e.g., 2 mg-C/L) of bulk and size-fractionated FA sample was added to the colloidal suspension along with the addition of the electrolytes, with a measurement procedure similar to that in the absence of FA samples. As for the electrophoretic mobility analysis, each sample was measured at least 3 times and data were presented as mean ± standard deviation. 2.5. Aggregation kinetics Time-resolved DLS was used to obtain the aggregation profiles of colloidal particles in the absence or presence of bulk and sizefractionated NOM by measuring the variations in hydrodynamic diameters over time (t). The aggregation rate constant, k, is determined from the relative rate of z-average hydrodynamic diameter changes with time (Chen et al., 2006; Chen and Elimelech, 2006):
k∞
1 dDh ðtÞ N0 dt t/0
(1)
where N0 is the initial particle concentration, and Dh(t) is the hydrodynamic diameter of colloidal particles at time t. The initial rate of increase in Dh(t) is obtained by determining the slope in the
linear regime during the initial 20 min. Since colloidal aggregation was highly dependent on the interaction energy, two aggregation regimes, that is, reaction and diffusion limited aggregation can be identified. Diffusion limited aggregation usually occurs when there is a low or negative energy barrier, under which the repulsive force is negligible and thus the aggregation rate is dependent solely on the time for clusters to encounter each other by diffusion. The reaction limited aggregation, on the other hand, occurs when there is a substantial but not insurmountable energy barrier, under which repulsive and attraction force were highly dependent upon the solution chemistry (e.g., ionic strength, electrolyte valence). Based on different aggregation regimes, the particle attachment efficiency (a) can be obtained to quantitatively compare the aggregation potential of colloidal particles under various conditions. It is defined as normalizing the measured k by those under diffusion limited condition, where the aggregation behavior is independent of the electrolyte concentration. As the concentrations of colloids in all aggregation experiments are the same, the value of a in the absence of NOM is calculated by the initial slope of the aggregation profile at a given electrolyte concentration normalized by kfast under the diffusion limited condition, whereas the a value in the presence of NOM was calculated by normalizing the early stage aggregation rate constants by kfast determined for the respective samples in the absence of NOM (Chen and Elimelech, 2007; Baalousha et al., 2013):
a¼
dDh ðtÞ dt
k ¼ kfast dDh ðtÞ dt
t/0
(2)
t/0; fast
where the kfast is the aggregation rate constant of colloidal particles under the diffusion limited condition. 2.6. Sorption of bulk and MW-fractionated NOM onto aquatic colloidal particles Batch sorption experiments were performed in duplicate to estimate the sorption capacity of bulk and MW-fractionated FA by NCPs and TICPs. Detailed procedures of the sorption experiment have been described elsewhere (Xu and Jiang, 2015) and given in the SI. 2.7. Other chemical analyses DOC concentrations in the bulk and MW-fractionated FA samples were measured by a TOC-VCPH analyzer (TOC-400, Shimadzu). UV absorption spectra were measured using a spectrophotometer (UV-2550, Shimadzu) over a wavelength range of 200e800 nm with 1 nm increments. Fluorescence EEM spectra were measured using a Cary Eclipse fluorescence spectrophotometer (Varian, USA). The spectra were gathered with subsequent scanning emission spectra from 250 to 550 nm at 2 nm increments by varying the excitation wavelength from 200 to 450 nm at 10 nm increments. The spectra were recorded at a scan rate of 1200 nm/min, using excitation and emission slit bandwidths of 5 nm. Observations on HR-TEM were conducted using a JEM 2100 field emission gun TEM (JOEL, Japan). The HR-TEM, selected area electron diffraction and energy dispersive X-ray (EDX) analyses were conducted using a 200 keV electron beam to characterize the sample appearance and structure. Powder XRD patterns were collected on a X-ray diffractometer (Siemens D5000) with Cu-Ka radiation 40 kV and 30 mA with a step size of 0.02 . In addition, the independent-sample T-test
H. Xu et al. / Water Research 142 (2018) 176e186
model was performed to examine the significance of differences between different sample sets. 3. Results and discussion 3.1. Characterization of aquatic colloids Fig. 1 shows typical TEM micrographs and EDX spectra for the colloids collected from different ecological regions of the lake. No significant difference in shapes was observed for NCPs between the two regions. They both contained a mixture of spherical, polygonal and elliptical-shaped colloids with an uneven size distribution, ranging from 20 nm to 200 nm. In addition, some organic-coated boundary can be found on the surface of NCPs, which would be attributed to the chemical and/or physical sorption of macromol€fer et al., 2007). The ecules (Buffle and Leppard, 1995; Scha conformation of TICPs, however, showed an origin-dependent feature. Specifically, TICPs from the macrophyte-dominant region were mostly polygonal, while those from algae-dominant lake waters were mainly spherical. Compared with the wider size range of NCPs, the individual TICPs from the two sampling regions contained mainly colloids with sizes between 20 and 40 nm (Fig. 1), showing a more uniform size distribution for the TICPs. The electron diffraction patterns demonstrated obvious crystalline properties, and chemical composition obtained from EDX analysis indicated that these TICPs contained similar major elements such as Si, Fe, Al, K, and Mg (Table S1 in the SI). Detailed analysis, however, showed that the TICPs from algae-dominant region exhibited higher mineral Al and Si contents than that from macrophytedominant region, which may be responsible for their different conformations. Results from XRD analysis further identified that the main mineral structure for the TICPs included quartz (SiO2), montmorillonite [Na0.3(Al, Mg)2Si4O10(OH)2$8H2O], nontronite [Na0.3Fe2Si4O10(OH)2$4H2O], and volkonskoite [Ca0.3(Cr, Mg)2(Si, Al)4O10(OH)2$4H2O] (Fig. S3 in the SI). These results indicated that Al-, Si-, and Fe-containing nanoparticles are the main mineralogical composition for the TICPs in lake waters. 3.2. Sources of colloidal particles in lake waters Fig. 2 compares the wind-induced concentrations of colloidal particles between natural waters and simulated experiments. The concentrations of TICPs in lake waters ranged from 6.3 ± 0.3 to 20.7 ± 2.2 mg/L and from 2.8 ± 0.2 to 11.4 ± 0.5 mg/L for the algaeand macrophyte-dominant regions, respectively, which were significantly higher than those detected in some deep lake waters (Filella et al., 2009). The concentrations of NCPs were about two times higher than those of TICPs in our samples, indicating the presence and importance of colloidal organic matter in the water column, which is consistent with previous observations for other freshwater systems (Zou et al., 2006; Stolpe et al., 2013; Cai et al., 2015). In addition, adsorption of macromolecules on inorganic colloid surfaces as well as the excretion of colloid-like organic substances from algae and macrophytes could also contribute to the high abundance of NCPs in natural waters (Xu et al., 2013; Xu and Jiang, 2015). Further analysis showed that concentrations of colloidal particles in both lake waters and simulated resuspension experiments increased with increasing wind speeds (Fig. 2a), demonstrating the important role of sediment re-suspension in the formation of colloids in shallow lake waters. In the algae-dominant region, as the wind speed increased from 2 to 8 m/s, the concentrations of NCPs and TICPs in simulated experiments increased from 12.5 ± 0.9 and 4.9 ± 0.3 mg/L to 30.2 ± 2.1 and 16.0 ± 1.1 mg/L, respectively. However, under the same wind speed conditions, the NCPs and TICPs
179
collected from lake waters varied from 23.1 ± 1.9 and 6.3 ± 0.3 mg/L to 58.4 ± 4.2 and 20.7 ± 2.2 mg/L, respectively. The differences in colloidal abundance between lake waters and simulated experiments demonstrated that 53.9 ± 1.8% of NCPs and 78.3 ± 1.1% of TICPs in lake waters could be derived from the sediment resuspension during windy conditions. The effect of sediment resuspension to the aquatic colloid abundance in macrophytedominant region (Fig. 2b) showed similar results that the contribution of sediment re-suspension to natural colloid abundance was 55.0 ± 3.5% for NCPs and 80.1 ± 2.6% for TICPs. The lower percentage contribution of sediment re-suspension for NCPs can be attributed to the presence of colloid-like organic substances in the water column and/or the adsorption of macromolecules on inorganic colloid surfaces.
3.3. Aggregation of aquatic colloids in electrolyte solutions without NOM Time-resolved aggregation profiles of aquatic colloids, including NCPs and TICPs isolated from lake waters, as a function of electrolyte concentrations are presented in Fig. S4 in the SI. Based on aggregation kinetics, the electrolyte-dependent attachment efficiencies can be obtained. As shown in Fig. 3, the aggregation behavior of NCPs and TICPs in both Naþ and Ca2þ solutions exhibited clearly a DLVO-type aggregation profile, suggesting that electrostatic interactions played an important role in stabilizing the aquatic colloids (Yin et al., 2015). This cation-induced aggregation behavior has also been reported for other bare and surfacemodified inorganic colloidal particles (Liu et al., 2011; Dong and Lo, 2013). In this study, the attachment efficiency exhibited an initial electrolyte-dependent enhancement followed by a nonsensitive stage with increasing electrolyte concentrations. The minimum electrolyte concentration required for the occurrence of this non-sensitivity was defined as critical coagulation concentration (CCC) (Chen et al., 2006). It was shown that the CCC values were highly dependent on the valence of electrolytes. Taken the TICPs as an example, the CCCs in Naþ-treatment were 18 ± 2 and 28 ± 2 mM for TICPs from the macrophyte- and algae-dominant regions, respectively, while those in Ca2þ-treatment decreased to 0.55 ± 0.05 and 0.50 ± 0.05 mM, respectively. The much lower CCC value of Ca2þ was ascribed to the high charge density and strong neutralization abilities of divalent electrolytes (Stankus et al., 2011). Since the CCC values of NaCl observed in this study were in the range of 18e28 mM for TICPs, those of CaCl2, calculated based on the Schulze-Hardy rule (Verwey et al., 1999), were thus within 0.28 and 0.44 mM, which were in the same range as the experimental data (Fig. 3). In addition to electrolyte valences, types of colloidal particles also played an important role in determining the CCC values. For example, the NCPs exhibited higher CCC values than TICPs both for the macrophyte- and algae-dominant lake waters. Specifically, the CCC values of Naþ and Ca2þ were 18e28 and 0.50e0.55 mM for TICPs, respectively, which were obviously lower than those for the NCPs (33e150 mM in Naþ and 0.75e1.60 mM in Ca2þ). Based on these CCCs, it is clear that NCPs were more stable or less sensitive to the effect of cations than TICPs in both mono- and divalent electrolyte solutions. The zeta-potentials of TICPs in algae- and macrophyte-dominant lake waters were 42.0 ± 0.3 and 38.5 ± 1.6 mV, respectively, while those of NCPs were in the range from 60.3 ± 1.8 to 60.9 ± 2.3 mV. Lower zeta-potentials and steric repulsion originated from the surface coating suggest that more electrolyte cations are needed to screen the negative charge for colloidal aggregation, which accounts for the higher CCC values of Naþ and Ca2þ for NCPs.
180
H. Xu et al. / Water Research 142 (2018) 176e186
Fig. 1. TEM micrograph of the NCPs (aeb) and TICPs (ced) from macrophyte- and algae-dominant lake waters, respectively. The electron diffraction patterns and EDX maps for the TICPs are shown at the bottom for chemical composition determination.
H. Xu et al. / Water Research 142 (2018) 176e186
181
Fig. 2. The re-suspension process of algae- (a) and macrophyte-dominant (b) lake sediments under different wind speeds and its contribution to the colloidal abundance in overlaying waters.
Fig. 3. Electrolyte-dependent variations in attachment efficiencies of colloidal particles (10 mg/L) collected from macrophyte-(a) and algae-(b) dominant lake waters. The critical coagulation concentration (CCC) values for each situation are marked for comparisons.
3.4. Aggregation kinetics of colloids in electrolyte solution with bulk and MW-fractionated NOM 3.4.1. Characterization of bulk and MW-fractionated NOM Information on MW distribution of FA was obtained via the application of ultrafiltration. Comparisons in DOC concentrations between the bulk and ultrafiltered samples showed that the FA tested was composed of 40% (in mass) of the <1 kDa LMW and 60% of the HMW-FA (1 kDa~0.45 mm). The percentage of the LMW fraction was similar to those observed in natural river NOM samples using the same membrane disc and MWCO (Xu and Guo, 2017) but much lower than that of the Western Siberia lakes (Shirokova et al., 2013). Spectroscopic data for comparing the inherent chemical properties between the bulk and MW-fractionated FA samples are shown in Fig. 4. The absorbance intensities of the >1 kDa HMW-FA were generally higher than those of bulk and LMW-FA at all wavelengths (Fig. 4a). This demonstrated that the aromatic components and hydrophobic structures were partitioned mainly in the HMW fraction, similar with other previous studies (Shen et al.,
2015; Yin et al., 2015). The fluorescent EEM spectra in the bulk FA sample detected four major peaks with Ex/Em maxima at 275/ (320e340), 230/(320e340), 275/(380e400) and 230/(380e400) nm representing peaks A, B, C, D, respectively (Fig. 4b), manifesting the abundance of carboxylic-like functional groups (Yan et al., 2013; Xu et al., 2016a). Among these peaks, peaks A and B with low aromaticity were substaintially located in the LMW-fraction (Fig. 4c), while the HMW-fraction with high aromaticity and hydrophobicity contained mainly peaks C and D (Fig. 4d), showing a MW-dependent feature in abundance and chemical composition in the bulk FA. Similar distribution patterns were also found in other natural waters (Xu and Guo, 2017). 3.4.2. NOM-molecular-size-dependent aggregation behavior of TICPs Based on the time-resolved aggregation profiles (Fig. S5-S8 in the SI), the variations in attachment efficiencies of TICPs as a function of different electrolyte concentrations in the presence of bulk and MW-fractioned FAs can be obtained (Fig. 5). Overall, the addition of FA matrix resulted in the increase in CCC values in both
182
H. Xu et al. / Water Research 142 (2018) 176e186
Fig. 4. Comparison in absorbance (a) and fluorescent (bed) spectra between the bulk and MW-fractionated FA on the same mass basis, 10 mg/L of DOC for each FA fraction.
mono- and divalent electrolytes, showing a FA-induced enhancement in colloidal stability (Dong and Lo, 2013). Specifically, due to the addition of FA matrix, the CCCs of Naþ shifted from 18 ± 2e28 ± 2 mM to 150 ± 5e170 ± 10 mM, and those of Ca2þ increased from 0.50 ± 0.05e0.55 ± 0.05 mM to 3.30 ± 0.05e4.20 ± 0.05 mM. The CCC values in the presence of bulk FA observed herein were comparable with previous studies showing a CCC of 170 mM for Naþ and 2.5 mM for Ca2þ for the aggregation of Al2O3 colloids in the presence of microbially derived extracellular polymeric substance (EPS) (Xu et al., 2016b). The increase in CCC values indicated that colloidal particles were prone to be more stable even with a low DOC concentration, e.g., ~2 mg-C/L. However, in some shallow lakes especially the eutrophic shallow lakes, DOC concentrations can reach up to about 20 mg-C/L (Xu and Jiang, 2013), which were significantly higher than those applied in the simulation study. Due to the negative charge imparted by organic matters, colloids in such high DOC environments are expected to be relatively stable. This mechanism could result in the persistence of high water turbidity in some eutrophic shallow lakes, such as Lakes Taihu, Chaohu, and Dianchi in China (Xing and Kong, 2007; Zhang et al., 2014). In addition to bulk FA, variations in attachment efficiencies of TICPs were also related to the inherent MW properties. For example, CCCs values with the addition of LMW-FA were 2.7e3.4 times higher for Naþ and 2.2e3.0 times higher for Ca2þ, respectively, compared to those without FA. In contrast, the CCCs in the presence of HMW-FA were 7.9e15.6 times higher for Naþ and
12.2e12.4 times higher for Ca2þ, respectively. These results clearly demonstrated that the HMW-FA enhanced colloidal stability much more efficiently than the LMW-FA. Similar with the aggregation profiles of TICPs, the NOM-induced stability enhancement has been reported for some model inorganic colloids or nanoparticles, such as nano zero-valent iron, titanium dioxide, nickel oxide, zinc oxide, silica colloidal particles (Zhang et al., 2009; Dong and Lo, 2013; Kteeba et al., 2017). However, the relationship between specific NOM MW ranges and colloidal stability enhancement has not been well quantified. Our findings here on MW-dependent aggregation heterogeneity in different electrolytes implied that previous observations of NOM-enhanced colloidal stability should mainly resulted from the HMW fraction of NOM matrix. To probe into mechanisms responsible for the stability enhancement, zeta potentials of TICPs in the presence of different electrolytes altered by the bulk and MW-fractionated FA were compared (Fig. S9 in the SI). As depicted, the zeta potential of TICPs was generally more negative in the presence of FA, even with the presence of divalent electrolytes. In addition, the zeta potential of TICPs in the presence of HMW-FA became even more negative compared to those in the presence of bulk or LMW-FA, except for the condition of high Naþ concentrations (Fig. S9 in the SI). The more negative surface charge indicated strong electrostatic repulsion, which accounts for the higher colloidal stability in the presence of HMW-FA. In addition, owing to the plenty of aromatic substances and hydrophobicity, the HMW-FA would be expected to have preferable adsorption on TICPs through hydrophobic and p-p
H. Xu et al. / Water Research 142 (2018) 176e186
183
Fig. 5. The attachment efficiency of TICPs from macrophyte- (a, b) and algae-dominant (c, d) lake waters in the presence of electrolyte solutions as a function of the bulk and MWfractionated FA. The attachment efficiency of TICPs without FA addition and the critical coagulation concentration (CCC) for each situation were also shown for comparisons. The concentrations of TICPs were 10 mg/L, while those of the bulk and MW-fractionated FA were 2 mg/L.
interactions (Yin et al., 2015). Our adsorption experiments also showed that HMW-FA exhibited higher adsorption intensities than the bulk and LMW-FA (Table S2 in the SI). High adsorptive ability of HMW-FA indicated the efficient formation of organic layers on the surface of TICPs, thus enhancing the zeta potential and colloidal stability via steric repulsion (Shen et al., 2015). Therefore, both the electrostatic repulsion and steric hindrance can account for the MW-dependent enhancement of colloidal stability for TICPs in lake waters. 3.4.3. Aggregation kinetics of NCPs affected by bulk and MWfractionated NOM Fig. 6 shows the MW-dependent aggregation behavior of NCPs in the presence of different electrolytes, with the detailed timeresolved aggregation profiles shown in Figs. S5-S8 in the SI. Compared with the electrolyte-induced aggregation profiles in the absence of FA, addition of bulk FA resulted in higher CCC values: 1.2e1.3 times higher for Naþ and 1.3e1.4 times higher for Ca2þ. As for the MW-fractionated FA, the LMW-FA increased the CCC values at 1.2e1.4 times higher for Naþ and 1.1e1.5 times higher for Ca2þ, respectively, and the addition of HMW-FA also caused a slight increase in CCC values at 1.4e1.5 times higher for Naþ and 1.3e1.5 times higher for Ca2þ, respectively. These results demonstrated that the increase extents of CCCs for NCPs were much less significant compared to those for TICPs both in mono- and divalent electrolytes due to the difference in colloidal composition. Yin et al. (2015) also showed that NOM enhanced the stability of bare silver nanoparticles but exhibited minor influence for the citrate-coated nanoparticles (~20 nm). However, another study reported that addition of NOM elevated the stability of both citrate- and polyvinylpyrrolidone-coated silver nanoparticles (~60 nm) in
solutions containing Naþ or low concentrations of Ca2þ, but inhibited the stability at high Ca2þ concentrations (Huynh and Chen, 2011). This indicated that the influences of NOM on the aggregation profiles of surface-coated colloidal particles also depended on other factors, such as colloid properties, coating agents, and specific electrolytes. Our results were consistent with those of Yin et al. (2015). To reveal the mechanism for the insensitivity of aggregation kinetics in response to FA addition, the surface charges and Fourier transform infrared (FTIR) spectra of NCPs under various conditions were measured. The zeta potential and FTIR spectra of NCPs in both mono- and divalent electrolytes was not significantly influenced by the bulk and MW-fractionated FA (Figs. S10 and S11 in the SI), demonstrating weak interactions between FA and the coated organic layers on NCP surfaces. The adsorption experiments also showed that the adsorption capacities of bulk and MW-fractionated FA on NCPs surfaces were generally lower (p < 0.05) than those on TICPs (Table S2 in the SI). Therefore, the presence of organic layers inhibited the further adsorption (or coating) of the bulk and MWfractionated FA on the surfaces of NCPs, which were responsible for the slight MW-dependent enhancement of CCC values. 3.5. Significance and implications Due to the high surface activity and potential influence on the toxicity and bio-availability of many contaminants, aquatic colloids have received increasing attention in recent years (Lead and Wilkinson, 2006; Chanudet and Filella, 2009; Filella et al., 2009; Louie et al., 2016). However, many previous studies focused mainly on the spatio-temporal distribution and abundance analysis, lacking the information on sources of the colloidal particles (Pokrovsky
184
H. Xu et al. / Water Research 142 (2018) 176e186
Fig. 6. Variations in attachment efficiencies of aquatic NCPs from macrophyte- (a, b) and algae-dominant (c, d) lake waters in the presence of bulk and MW-fractionated FA. The attachment efficiency of NCPs without FA and the critical coagulation concentration (CCC) for each situation were also provided for comparisons. The concentrations of NCPs were 10 mg/L, while those of the bulk and MW-fractionated FA were 2 mg/L.
et al., 2016). In this study, a sediment generation apparatus was applied to quantify the contribution of sediment re-suspension to the colloid abundance in natural waters. It was found that, under environmentally relevant conditions, the process of sediment resuspension contributed up to 78e80% of TICPs and 54e55% of NCPs in the water column of eutrophic shallow lakes. To our knowledge, this is the first study elucidating the sources and production of natural colloids. As a ubiquitous entity, NOM can regulate the stability of colloids via electrostatic repulsion and steric hindrance (Furman et al., 2013; Philippe and Schaumann, 2014). Although previous studies on the NOM effects on colloidal stability were mainly based on model inorganic colloids, a consistent NOM-induced stability enhancement has been observed in both mono- and divalent electrolytes, similar to those observed for TICPs here. However, we found that the NOM-induced aggregation heterogeneities of the NCPs were less significant than those of the TICPs. As known, engineered nanoparticles are likely to be coated with organic macromolecules, either intentionally added by manufacturers for desired properties or incidentally obtained during their life cycle (Louie et al., 2013). Moreover, the large specific surface area endowed inorganic colloids a high capacity for NOM adsorption or coating (Table S2 in the SI), forming the predominant NCPs species in aquatic environments. Thus, due to the noticeable heterogeneities in aggregation profiles between the NCPs and TICPs, previous results on colloidal behavior analysis using model inorganic colloids are likely to deviate from those of aquatic colloids and may need a critical reevaluation. Furthermore, since the bulk NOM is highly heterogeneous in molecular size and composition, different molecular size fractions
of NOM are expected to play distinct roles in determining the aggregation behavior. Compared with previous studies aiming at the effects of bulk NOM (Zhang et al., 2009; Dong and Lo, 2013), we found evident MW-dependent aggregation heterogeneity in both mono- and divalent electrolytes with the >1 kDa HMW-FA having higher stability efficiency, followed by bulk FA and then the <1 kDa LMW-FA. This is consistent with the distribution of aromatic and hydrophobic organic components among the LMW-, HMW- and bulk NOM pools (Cai et al., 2015). The results highlighted the heterogeneity of NOM in regulating colloidal stability among different MW fractions within NOM matrix. Overall, our results revealed that even a low concentration of organic ligands (~2 mg-C/L) can significantly enhance the colloidal stability, and divalent cations enhanced colloidal stability more efficiently than monovalent cations. As reported, fresh waters usually contain a Ca2þ concentration less than 1.0 mM and Naþ less than 2.5 mM (Xu et al., 2016b). This indicated that the colloidal particles, including NCPs and TICPs, in most freshwater systems, were usually characterized as obvious reaction-limited aggregation regime. This also accounted for the persistence of high water turbidity in many aquatic ecosystems. The mechanisms for colloidal aggregation in natural waters included electrostatic neutralization and/or interparticle bridging (Huangfu et al., 2013; Philippe and Schaumann, 2014). Our previous studies reported that, in the presence of cyanobacterial EPS, electrostatic neutralization contributed to the aggregation of model inorganic colloids at low Ca2þ concentration, while cation bridging became the main mechanism for the enhanced aggregation in Ca2þ concentrations higher than the CCC (Xu et al., 2016b). Yin et al. (2015) also found that although HMW-NOM was the key
H. Xu et al. / Water Research 142 (2018) 176e186
component in stabilizing the citrate-coated silver colloids at low Ca2þ concentrations, it could play a reversed role in promoting the colloidal aggregation at high concentrations of divalent cations. However, the phenomenon of the interparticle bridging was not observed in this study both for the NCPs and TICPs. The reason for the absence of bridging mechanism was still unclear, but may be related to factors, such as pH, NOM composition, molecular size, and colloidal structure, which needs further studies. 4. Conclusion Aquatic colloids, including NCPs and TICPs, in eutrophic shallow lake waters, were characterized with high abundance, broad size distribution, and Fe/Al/Si-containing mineralogical properties. Sediment re-suspension can have a significant contribution to colloidal particles at 54e55% of the NCPs and 78e80% of the TICPs in lake waters. The colloidal particles exhibited obvious DLVO-like aggregation profiles in mono- and divalent electrolytes, but presence of FA significantly inhibited the aggregation and enhanced the colloidal stability in the water column. Different MW fractions within FA matrix showed different effects on colloidal stability against aggregation with HMW fraction exhibiting higher stability efficiency than the LMW counterpart. However, the MWdependent aggregation heterogeneities for NCPs were less significant than those for TICPs. Therefore, in addition to bulk NOM, influence of DOM with different size fractions within NOM matrix should also be considered when interpreting effects of NOM on colloidal behavior. In addition differences in the NOM MWdependent aggregation behavior between NCPs and TICPs indicated that previous results on colloidal behavior analysis using model inorganic colloids should be reevaluated. Acknowledgements We gratefully thank three anonymous reviewers for their constructive comments and suggestions which improved the manuscript. This study was supported, in part, by the National Natural Science Foundation of China (51479187) and Youth Innovation Promotion Association CAS (2016286). We also thank Dr. Cheng Liu and Qiushi Shen for their help in sediment re-suspension simulation. Appendix A. Supplementary data Supplementary data related to this article can be found at https://doi.org/10.1016/j.watres.2018.05.059. References Alasonati, E., Slaveykova, V.I., Gallard, H., Croue, J.P., Benedetti, M.F., 2010. Characterization of the colloidal organic matter from the Amazonian basin by asymmetrical flow field-flow fractionation and size exclusion chromatography. Water Res. 44 (1), 223e231. Aurell, C.A., Wistrom, A.O., 2000. Coagulation of kaolinite colloids in high carbonate strength water. Colloid. Surface. Physicochem. Eng. Aspect. 168 (3), 277e285. Baalousha, M., Nur, Y., Romer, I., Tejamaya, M., Lead, J.R., 2013. Effect of monovalent and divalent cations, anions and fulvic acid on aggregation of citrate-coated silver nanoparticles. Sci. Total Environ. 454, 119e131. Borgnino, L., 2013. Experimental determination of the colloidal stability of Fe(III)montmorillonite: effects of organic matter, ionic strength and pH conditions. Colloid. Surface. Physicochem. Eng. Aspect. 423, 178e187. Buffle, J., Leppard, G.G., 1995. Characterization of aquatic colloids and macromolecules. 1. structure and behavior of colloidal material. Environ. Sci. Technol. 29 (9), 2169e2175. Cai, Y.H., Guo, L.D., Wang, X.R., Aiken, G., 2015. Abundance, stable isotopic composition, and export fluxes of DOC, POC, and DIC from the Lower Mississippi River during 2006e2008. J. Geophys. Res.: Biogeosci. 120 (11), 2273e2288. Chanudet, V., Filella, M., 2007. The fate of inorganic colloidal particles in Lake Brienz. Aquatic Sci.-Res. Across Boundaries 69 (2), 199e211.
185
Chanudet, V., Filella, M., 2009. Size and composition of inorganic colloids in a perialpine, glacial flour-rich lake. Geochem. Cosmochim. Acta 72 (5), 1466e1479. Chen, K.L., Elimelech, M., 2006. Aggregation and deposition kinetics of fullerene (C60) nanoparticles. Langmuir 22 (26), 10994e11001, 2006. Chen, K.L., Mylon, S.E., Elimelech, M., 2006. Aggregation kinetics of alginate-coated hematite nanoparticles in monovalent and divalent electrolytes. Environ. Sci. Technol. 40 (5), 1516e1523. Chen, K.L., Elimelech, M., 2007. Influence of humic acid on the aggregation kinetics of fullerene (C60) nanoparticles in monovalent and divalent electrolyte solutions. J. Colloid Interface Sci. 309 (1), 126e134. Colman, B.P., Espinasse, B., Richardson, C.J., Matson, C.W., Lowry, G.V., Hunt, D.E., Wiesner, M.R., Bernhardt, E.S., 2014. Emerging contaminant or an old toxin in disguise? Silver nanoparticle impacts on ecosystems. Environ. Sci. Technol. 48 (9), 5229e5236. Dong, H.R., Lo, I.M.C., 2013. Influence of humic acid on the colloidal stability of surface-modified nano zero-valent iron. Water Res. 47 (1), 419e427. Dumat, C., Cheshire, M.V., Fraser, A.R., Shand, C.A., Staunton, S., 1997. The effect of removal of soil organic matter and iron on the adsorption of radiocaesium. Eur. J. Soil Sci. 48 (4), 675e683. Filella, M., Chanudet, V., Philippo, S., Quentel, F., 2009. Particle size and mineralogical composition of inorganic colloids in waters draining the adit of an abandoned mine, Goesdorf, Luxembourg. Appl. Geochem. 24 (1), 52e61. Furman, O., Usenko, S., Lau, B.L.T., 2013. Relative importance of the humic and fulvic fractions of natural organic matter in the aggregation and deposition of silver nanoparticles. Environ. Sci. Technol. 47 (3), 1349e1356. Garcia-Garcia, S., Wold, S., Jonsson, M., 2007. Kinetic determination of critical coagulation concentrations for sodium- and calcium-montmorillonite colloids in NaCl and CaCl2 aqueous solutions. J. Colloid Interface Sci. 315 (2), 512e519. Gibson, C.T., Turner, I.J., Roberts, C.J., Lead, J.R., 2007. Quantifying the dimensions of nanoscale organic surface layers in natural waters. Environ. Sci. Technol. 41 (4), 1339e1344. guen, C., Guo, L.D., Wang, D.L., Tanaka, N., Hung, C.C., 2006. Chemical characGue teristics and origin of dissolved organic matter in the Yukon River. Biogeochemistry 77 (2), 139e155. Guo, L.D., Santschi, P.H., Ray, S.M., 2002. Metal partitioning between colloidal and dissolved phases and its relation with bioavailability to American oysters. Mar. Environ. Res. 54 (1), 49e64. Guo, L.D., White, D.M., Xu, C., Santschi, P.H., 2009. Chemical and isotopic composition of HMW-DOM from the Mississippi River plume. Mar. Chem. 114, 63e71. Huangfu, X.L., Jiang, J., Ma, J., Liu, Y.Z., Yang, J., 2013. Aggregation kinetics of manganese dioxide colloids in aqueous solution: influence of humic substances and biomacromolecules. Environ. Sci. Technol. 47 (18), 10285e10292. Huynh, K.A., Chen, K.L., 2011. Aggregation kinetics of citrate and polyvinylpyrrolidone coated silver nanoparticles in monovalent and divalent electrolyte solutions. Environ. Sci. Technol. 45 (13), 5564e5571. Jiang, C.L., Sequaris, J.M., Vereecken, H., Klumpp, E., 2012. Effects of inorganic and organic anions on the stability of illite and quartz soil colloids in Na-, Ca- and mixed Na-Ca systems. Colloid. Surface. Physicochem. Eng. Aspect. 415, 134e141. Kretzschmar, R., Holthoff, H., Sticher, H., 1998. Influence of pH and humic acid on coagulation kinetics of kaolinite: a dynamic light scattering study. J. Colloid Interface Sci. 202 (1), 95e103. Kteeba, S.M., El-Adawi, H.I., El-Rayis, O.A., El-Ghobashy, A.E., Schuld, J.L., Svoboda, K.R., Guo, L.D., 2017. Zinc oxide nanoparticle toxicity in embryonic zebrafish: mitigation with different natural organic matter. Environ. Pollut. 230, 1125e1140. Lead, J.R., Wilkinson, K.J., 2006. Aquatic colloids and nanoparticles: current knowledge and future trends. Environ. Chem. 3 (3), 159e171. Liu, C., Shen, Q.S., Zhou, Q.L., Fan, C.X., Shao, S.G., 2015. Precontrol of algae-induced black blooms through sediment dredging at appropriate depth in a typical eutrophic shallow lake. Ecol. Eng. 77, 139e145. Liu, X.Y., Wazne, M., Chou, T.M., Xiao, R., Xu, S.Y., 2011. Influence of Ca2þ and Suwannee river humic acid on aggregation of silicon nanoparticles in aqueous media. Water Res. 45 (1), 105e112. Louie, S.M., Tilton, R.D., Lowry, G.V., 2013. Effects of molecular weight distribution and chemical properties of natural organic matter on gold nanoparticle aggregation. Environ. Sci. Technol. 47 (9), 4245e4254. Louie, S.M., Spielman-Sun, E.R., Small, M.J., Tilton, R.D., Lowry, G.V., 2015. Correlation of the physicochemical properties of natural organic matter samples from different sources to their effects on gold nanoparticle aggregation in monovalent electrolyte. Environ. Sci. Technol. 49 (4), 2188e2198. Louie, S.M., Tilton, R.D., Lowry, G.V., 2016. Critical review: impacts of macromolecular coatings on critical physicochemical processes controlling environmental fate of nanomaterials. Environmental Science: Nano 3 (2), 283e310. Philippe, A., Schaumann, G.E., 2014. Interactions of dissolved organic matter with natural and engineered inorganic colloids: a review. Environ. Sci. Technol. 48 (16), 8946e8962. Pokrovsky, O.S., Manasypov, R.M., Loiko, S.V., Shirokova, L.S., 2016. Organic and organo-mineral colloids in discontinuous permafrost zone. Geochem. Cosmochim. Acta 188, 1e20. Qin, B.Q., Xu, P.Z., Wu, Q.L., Luo, L.C., Zhang, Y.L., 2007. Environmental issues of lake Taihu, China. Hydrobiologia 581 (1), 3e14. €fer, T., Chanudet, V., Claret, F., Filella, M., 2007. Spectromicroscopy mapping of Scha colloidal/particulate organic matter in Lake Brienz, Switzerland. Environ. Sci. Technol. 41 (22), 7864e7869. Sequaris, J.M., 2010. Modeling the effects of Ca2þ and clay-associated organic carbon
186
H. Xu et al. / Water Research 142 (2018) 176e186
on the stability of colloids from topsoils. J. Colloid Interface Sci. 343 (2), 408e414. Shang, E.X., Li, Y., Niu, J.F., Zhou, Y.J., Wang, T.Y., Crittenden, J.C., 2017. Relative importance of humic and fulvic acid on ROS generation, dissolution, and toxicity of sulfide nanoparticles. Water Res. 124, 595e604. Shen, M.H., Yin, Y.G., Booth, A., Liu, J.F., 2015. Effects of molecular weight-dependent physicochemical heterogeneity of natural organic matter on the aggregation of fullerene nanoparticles in mono-and di-valent electrolyte solutions. Water Res. 71, 11e20. Shirokova, L.S., Pokrovsky, O.S., Kirpotin, S.N., Desmukh, C., Pokrovsky, B.G., Audry, S., Viers, J., 2013. Biogeochemistry of organic carbon, CO2, CH4, and trace elements in thermokarst water bodies in discontinuous permafrost zones of Western Siberia. Biogeochemistry 113, 573e593. Stankus, D.P., Lohse, S.E., Hutchison, J.E., Nason, J.A., 2011. Interactions between natural organic matter and gold nanoparticles stabilized with different organic capping agents. Environ. Sci. Technol. 45 (8), 3238e3244. Stolpe, B., Guo, L.D., Shiller, A.M., Aiken, G.R., 2013. Abundance, size distributions and trace-element binding of organic and iron-rich nanocolloids in Alaskan rivers, as revealed by field-flow fractionation and ICP-MS. Geochem. Cosmochim. Acta 105, 221e239. Verwey, E.J.W., Overbeek, J.T.G., Overbeek, J.T.G., 1999. Theory of the Stability of Lyophobic Colloids. Courier Corporation. Xing, P., Kong, F.X., 2007. Intra-habitat heterogeneity of environmental factors regulating bacterioplankton community composition in Lake Taihu, China. Aquat. Microb. Ecol. 48 (2), 113e122. Xu, H.C., Jiang, H.L., 2013. UV-induced photochemical heterogeneity of dissolved and attached organic matter associated with cyanobacterial blooms in a eutrophic freshwater lake. Water Res. 47 (17), 6506e6515. Xu, H.C., Cai, H.Y., Yu, G.H., Jiang, H.L., 2013. Insights into extracellular polymeric substances of cyanobacterium Microcystis aeruginosa using fractionation procedure and parallel factor analysis. Water Res. 47 (6), 2005e2014. Xu, H.C., Jiang, H.L., 2015. Effects of cyanobacterial extracellular polymeric substances on the stability of ZnO nanoparticles in eutrophic shallow lakes.
Environ. Pollut. 197, 231e239. Xu, H.C., Lv, H., Liu, X., Wang, P.F., Jiang, H.L., 2016a. Electrolyte cations binding with extracellular polymeric substances enhanced Microcystis aggregation: implication for Microcystis bloom formation in eutrophic freshwater lakes. Environ. Sci. Technol. 50 (17), 9034e9043. Xu, H.C., Yang, C.M., Jiang, H.L., 2016b. Aggregation kinetics of inorganic colloids in eutrophic shallow lakes: influence of cyanobacterial extracellular polymeric substances and electrolyte cations. Water Res. 106, 344e351. Xu, H.C., Guo, L.D., 2017. Molecular size-dependent abundance and composition of dissolved organic matter in river, lake and sea waters. Water Res. 117, 115e126. Yan, M.Q., Fu, Q.W., Li, D.C., Gao, G.F., Wang, D.S., 2013. Study of the pH influence on the optical properties of dissolved organic matter using fluorescence excitationeemission matrix and parallel factor analysis. J. Lumin. 142, 103e109. Yan, Z.S., Jiang, H.L., Li, X.H., Shi, Y., 2014. Accelerated removal of pyrene and benzo [a]pyrene in freshwatersediments with amendment of cyanobacteria-derived organic matter. J. Hazard Mater. 272, 66e74. Yin, Y.G., Shen, M.H., Tan, Z.Q., Yu, S.J., Liu, J.F., Jiang, G.B., 2015. Particle coatingdependent interaction of molecular weight fractionated natural organic matter: impacts on the aggregation of silver nanoparticles. Environ. Sci. Technol. 49 (11), 6581e6589. Zhang, Y., Chen, Y.S., Westerhoff, P., Crittenden, J., 2009. Impact of natural organic matter and divalent cations on the stability of aqueous nanoparticles. Water Res. 43 (17), 4249e4257. Zhang, Y.L., Shi, K., Liu, X.H., Zhou, Y.Q., Qin, B.Q., 2014. Lake topography and wind waves determining seasonal-spatial dynamics of total suspended matter in turbid Lake Taihu, China: assessment using long-term high-resolution MERIS data. PLoS One 9 (5), e98055. Zheng, S.S., Wang, P.F., Wang, C., Hou, J., 2015. Sediment resuspension under action of wind in Taihu Lake, China. Int. J. Sediment Res. 30 (1), 48e62. Zou, L., Sun, M.Y., Guo, L.D., 2006. Temporal variations of organic carbon inputs into the upper Yukon River: evidence from fatty acids and their stable carbon isotopic compositions in dissolved, colloidal and particulate phases. Org. Geochem. 37 (8), 944e956.