Assessing the Effects of Silver Nanoparticles on ... - ACS Publications


Dec 23, 2013 - Vincent C. ReyesPhillip B. GedalangaNancy MerinoJoy D. Van .... Hannele Auvinen , Vincent Gagnon , Diederik P. L. Rousseau , Gijs Du La...
0 downloads 0 Views 2MB Size


Article pubs.acs.org/est

Assessing the Effects of Silver Nanoparticles on Biological Nutrient Removal in Bench-Scale Activated Sludge Sequencing Batch Reactors Christina L. Alito and Claudia K. Gunsch* Department of Civil and Environmental Engineering, Duke University, Box 90287, Durham, North Carolina 27708, United States Center for Environmental Implications of NanoTechnology (CEINT), Duke University, Durham, North Carolina 27708, United States S Supporting Information *

ABSTRACT: Consumer products such as clothing and medical products are increasingly integrating silver and silver nanoparticles (AgNPs) into base materials to serve as an antimicrobial agent. Thus, it is critical to assess the effects of AgNPs on wastewater microorganisms essential to biological nutrient removal. In the present study, pulse and continuous additions of 0.2 and 2 ppm gum arabic and citrate coated AgNPs as well as Ag as AgNO3 were fed into sequencing batch reactors (SBRs) inoculated with nitrifying sludge. Treatment efficiency (chemical oxygen demand (COD) and ammonia removal), Ag dissolution measurements, and 16S rRNA bacterial community analyses (terminal restriction fragment length polymorphism, T-RFLP) were performed to evaluate the response of the SBRs to Ag addition. Results suggest that the AgNPs may have been precipitating in the SBRs. While COD and ammonia removal decreased by as much as 30% or greater directly after spikes, SBRs were able to recover within 24 h (3 hydraulic retention times (HRTs)) and resume removal near 95%. T-RFLP results indicate Ag spiked SBRs were similar in a 16s rRNA bacterial community. The results shown in this study indicate that wastewater treatment could be impacted by Ag and AgNPs in the short term but the amount of treatment disruption will depend on the magnitude of influent Ag.



INTRODUCTION Ionic silver is an established antibacterial agent and has been utilized for its antiseptic properties in consumer products such as washing machines, clothing, and children’s toys.1 More recently, specially coated silver nanoparticles (AgNPs) have been designed to replace ionic silver in consumer products because of their ability to slowly release ionic silver as their coatings dissolve, making them better suited for long-term consumer product usage. Because manufactured AgNPs generally range from 10 to 100 nm, researchers speculate that they may be small enough to pass through the cell membrane of a microorganism and, then, release ionic silver directly into the microbe as their coatings dissolve.2 Since it is likely that some consumer products such as textiles will be washed, it is critical that the concentration of AgNPs in wastewater treatment plant (WWTP) influent be quantified and their impacts ascertained. Current reports show that there is a high variability in silver concentration present in consumer products and it is difficult to find estimations of the amount of total silver and AgNPs being added to consumer products. Benn et al.3 found that AgNP containing socks leached up to 68 μg silver/g sock in water with gentle agitation, while Geranio et al.4 measured up to 377 μg/g with the application of detergent. Since AgNP-containing consumer products are already in use, several methods have been developed for estimating © 2013 American Chemical Society

speciation of AgNPs in complex media such as wastewater, where it will likely collect. Because one of the suggested antimicrobial mechanisms of AgNPs is the release of ionic silver after the particle coating dissociates, researchers are examining nanoparticle coating interactions with solution properties such as pH, ionic strength, and dissolved organic carbon concentration.5 The characteristically high concentration of thiols in wastewater sludge has also led to the discovery of naturally occurring silver sulfide nanoparticles in wastewater, which provides evidence that influent silver is likely to be bound up with wastewater sludge thiols.6 Nonetheless, because of their proven antimicrobial properties, even if AgNPs are only present at very low concentrations in wastewater influent, it is possible that AgNPs could alter important microbial functions. The disruption of biological nutrient removal, and especially nitrification, in wastewater treatment is of particular concern. Nitrifying bacteria including Nitrosomonas spp. can be easily inactivated by disruption of their membrane-bound enzyme ammonia monooxygenase (AMO) which could ultimately cause nitrification treatment failures.7 Received: Revised: Accepted: Published: 970

August 15, 2013 December 20, 2013 December 23, 2013 December 23, 2013 dx.doi.org/10.1021/es403640j | Environ. Sci. Technol. 2014, 48, 970−976

Environmental Science & Technology

Article

between 6 and 6.5 mg/L. The reactors were continuously mixed with large stir bars on stir plates at 700 rpm. The cycling of air, feed, and pumping was controlled by Intermatic TN311C timer-controllers (Grove, IL). Reactors were kept at room temperature (∼17−20 °C). The SBRs were initially spiked with 500 mL of nitrifying activated sludge taken from the North Durham Water Reclamation Facility (Durham, NC). This plant currently treats up to 20 MGD and utilizes biological nutrient removal systems to remove phosphorus, BOD, and ammonia. Prior to beginning the AgNP spikes, the SBRs were operated for 120 d with no Ag additions. It took the reactors 90 d to reach steady state. The reactors were allowed to perform at steady state for 30 d prior to any AgNP addition; they were steadily removing 90% or more of COD and ammonia during this period. Microbial community structure analysis was also performed on steady state SBRs to determine the starting community diversity. Reactors were divided into 2 replicate groups of 4 reactors based on the similarity of their starting microbial community profiles. In total, four SBR treatments were conducted in duplicate consisting of: (1) no Ag control, (2) citrate AgNP treatment, (3) GA AgNP treatment, and (4) Ag+ control. At the time of the initial Ag spike, the COD and ammonia removal rates were not significantly different across all eight reactors (p > 0.05). AgNP Addition. Concentrations of 0.2 and 2 ppm were selected as the low and high Ag additions. These concentrations were chosen on the basis of our previous study, which showed that the effects of AgNPs on the model wastewater ammonium oxidizing bacterium (AOB), Nitrosomonas europaea, became apparent around 2 ppm total Ag concentration.10 Table 1

The bactericidal nature of silver has been well established in previous literature,8 but the toxicity of AgNPs appears to differ greatly depending on size and coating.9 Due to increased utilization of AgNPs containing consumer products, further evaluation is needed to determine potential impacts of AgNPs. Although a handful of studies have examined the antimicrobial effects of AgNPs on wastewater bacteria, none have examined the effects on wastewater bacterial diversity and overall treatment efficiency by cycling pulse and continuous inputs of Ag. The objective of the present study was to determine the effect of two AgNPs (citrate and gum arabic stabilized) on common treatment characteristics (namely, in terms of chemical oxygen demand (COD) and ammonia removal efficiency) and microbial community structure in sequencing batch reactors (SBRs) mimicking WWTP operation. These types of AgNPs were chosen to follow up our previous research, which studied the effects of AgNPs on the nitrifier, Nitrosomonas europaea.10 Gum arabic and citrate stabilized AgNPs were found to be to most toxic to the sensitive nitrifiers responsible for converting ammonia to nitrite.



MATERIALS AND METHODS AgNP Characterization. Gum arabic (GA) and citrate stabilized AgNPs were used in the present study. These particles were selected because they have been widely studied and because their coatings are representative of AgNPs used in consumer products, as previously described.10 Citrate AgNPs were produced by reducing Ag nitrate in water with sodium citrate.11 GA AgNPs were manufactured by reducing Ag nitrite with water and GA. The average particle sizes measured by transmission electron microscopy (TEM) were 32.3 ± 0.5 and 15.5 ± 0.5 nm for GA and citrate AgNPs, respectively. More information related to AgNP synthesis and characterization can be found in Yin et al.12 and Meyer et al.13 In addition, the “AgNP Synthesis” section in the Supporting Information of Arnaout and Gunsch10 has extensive information on TEM images, zeta sizing of AgNPs, and laboratory procedures. SBR Design. Eight 3-L bench-scale SBRs were constructed out of square pieces of plexiglass (6 × 6 × 6 in, 1/8 in. thick), which were welded together using poly(methyl methacrylate). The peak volume in all reactors was 2.0 L, and the volume after decanting was approximately 1 L. SBRs were operated on an 8 h cycle, starting with 30 min of influent synthetic wastewater feeding, 6 h of aerating and vigorous mixing, 1 h of settling, and 30 min of decanting. A 13 d solids retention time (SRT) was chosen to ensure optimum conditions for nitrification, and the reactors were run with a 5.3 h hydraulic retention time (HRT), which falls in the typical range for contact stabilization systems.14 The SRT was maintained by wasting sludge periodically to retain 2000−2500 mg/L of mixed liquor suspended solids (MLSS). The SBRs were fed synthetic wastewater (SWW) based on a previously published recipe,15 which had an average COD of 450 mg/L, average ammonia concentration of 40 mg/L, and pH of 7. All reactors were covered with aluminum foil to prevent any photolysis. AgNPs were added in stock concentration either directly to the reactors (for pulse additions) or into the influent medium (for continuous additions). Figure S1, Supporting Information, shows the configuration of SBRs built in this study. Air was fed into reactors via Tetra Whisper small aquatic aerators (Madison, WI), and feedwater was pumped in via multihead pumps (Cole Parmer, Masterflex L/S, Vernon Hills, IL). The dissolved oxygen was maintained

Table 1. Order of Ag Addition to SBRsa order of addition

spike concentration, ppm

type of spike

length of acclimation cycle

1 2 3 4

0.2 0.2 2 2

pulse continuous pulse continuous

14 d, 6 d, 2 d 3 SRTs 14 d, 6 d, 2 d 3 SRTs

a SBRs were first spiked with 14 d, 6 d, or 2 d pulse Ag additions, followed by a phase of continuous Ag addition. After the first 0.2 ppm Ag cycle, the concentration was increased to 2 ppm and the cycling was restarted.

displays the chronology of Ag addition to the SBRs. Ag was initially added in pulses at 0.2 ppm followed by a continuous addition. For the pulse input, either AgNPs or AgNO3 was added directly to the SBRs as a single pulse at the start of the cycles. Following each pulse input, the SBRs were allowed to adapt with no Ag addition for different lengths of time. The first acclimation cycle was the longest (14 d) to assess the recovery period needed by the SBRs, and then, shorter acclimation cycles followed (6 and 2 d, respectively). Following the pulse inputs, a continuous flow of Ag was added by spiking influent SWW with stock solutions of AgNPs and AgNO3. The concentration of Ag was increased to 2 ppm at the end of the 0.2 ppm continuous phase, and the pulse cycling was restarted (14 d pulse, 6 d pulse, 2 d pulse, continuous). Analytical Methods. To monitor COD and N removal efficiency, influent and effluent samples were collected approximately every 3 d as well as immediately after each Ag spike. Four parameters were measured using HACH reagents (Loveland, CO): COD (mercuric digestion method), ammonia 971

dx.doi.org/10.1021/es403640j | Environ. Sci. Technol. 2014, 48, 970−976

Environmental Science & Technology

Article

gradually decreased after 24 h. Dissolved Ag remained below 4 ppb for all 0.2 and 2 ppm total Ag spikes (Table S1, Supporting Information). Figure 1 shows the drop in total Ag

(salicylate method), nitrite (diazotization method), and nitrate (cadmium reduction method). Samples were taken directly from SBR effluent containers and tested within 30 min. To measure the Ag concentration in the SBRs, samples were collected directly from the SBRs during treatment as well as from the treated SBR effluent. Samples for Ag spikes were taken at 1 min, 1 h, 6 h, 24 h, and at the end of the spike, while samples for continuous Ag spikes were taken at 24 h and every 7 d following the continuous spike. Effluent treated water samples were also collected for Ag analysis every 3 d. Ag analysis was performed as previously described in Arnaout and Gunsch10 and in the Supporting Information section. DNA Extraction and PCR Conditions. Biomass samples were taken periodically from the SBRs and microcentrifuged for 1 min at 13 000g. Biomass samples were then immediately stored at −20 °C until DNA extraction. Duplicate samples were extracted for each treatment using the UltraClean DNA Isolation Kit (MoBio Laboratories, Solana Beach, CA). PCR of the bacterial 16S SSU rRNA gene region was carried out by the methods described in Lukow et al.16 with small adjustments. 6-Carboxyfluorescein (6-FAM) was used to fluorescently label the forward primer (27F), and 1392R was used as the unlabeled reverse primer.17 Please refer to the Supporting Information section for a further description of DNA extraction. Terminal Restriction Fragment Length Polymorphism (T-RFLP) Analysis. Restriction enzyme digests were carried out by following the protocol described in Lukow et al.16 MspI was added to purified PCR product and incubated for 3 h at 37 °C. Fragment analysis was carried out using POP6 polymer and ROX-labeled MapMarker 1000 size standards. Applied Biosystems GeneScan v3.7.1 software (Foster City, CA) was used to interpret the raw data after electrophoresis. All samples were visually inspected for clean peaks, and raw data was transferred into T-REX.18 T-REX aligned peaks and identified true peaks. Data analysis was performed in T-REX using the peak area feature, so that abundance and diversity could be examined. T-RFs smaller than 50 bp were not included in analysis, as to eliminate primer dimer fragments. Data files were then imported into PAST statistical software to create Bray− Curtis principle coordinate ordination and cluster similarity (Hammer & Harper, D.A.T. 2006. Paleontological Data Analysis. Blackwell). Diversity indices were calculated by comparing taxonomic species with the Simpson 1-D index. The greater the resulting value, the more diverse is the sample in species. Please refer to the Supporting Information section for a detailed description of T-RFLP methodology. Statistical Analysis. The unpaired, two tailed student’s t test was used to identify statistical differences between control samples and treated samples for all data analysis except T-RFLP microbial community statistical analysis. Bray−Curtis similarity and principle coordinate ordinations were utilized for T-RFLP data interpretation.

Figure 1. Total Ag concentration in SBRs during pulse and continuous inputs. (A) 2 ppm spikes and (B) 0.2 ppm spikes. Bar diagrams show the rapid decrease in total Ag measured from the supernatant immediately after pulse spikes and during the continuous phase of Ag addition.

concentration during the 14 d spike and the average concentration during the continuous input of Ag. For the 0.2 ppm spike, the total Ag concentrations in the wastewater after 1 h for the 14 d pulse spike were 96.8 ± 0.95, 70.8 ± 10.9, and 134.2 ± 10.8 ppb Ag for citrate AgNPs, GA AgNPs, and Ag as AgNO3, respectively. The dissolved Ag concentrations were not statistically significantly different when compared to the no Ag control concentrations (p > 0.05). Total and dissolved Ag concentrations in the no Ag control SBR were consistently lower than ICP-MS detection limit ( 0.05). The average total concentrations over 30 d during the continuous addition were 9.28 ± 10.9, 6.94 ± 4.2, and 9.37 ± 12.9 ppb for citrate AgNPs, GA AgNPs, and Ag as AgNO3, respectively. Dissolved Ag concentrations were below detection. Similarly to



RESULTS AND DISCUSSION Dissolution of Ag from AgNPs. The concentration of total and dissolved Ag was measured in the SBRs to assess how quickly Ag dissolved from the AgNPs. Since Ag was measured in the influent, effluent, and the supernatant of the SBRs, the partition of the Ag in the SBRs could be deduced from Ag concentration measurements. In general, total Ag concentrations during pulse inputs were near dosing concentration (0.2 and 2 ppm) immediately following spikes (1 min) and 972

dx.doi.org/10.1021/es403640j | Environ. Sci. Technol. 2014, 48, 970−976

Environmental Science & Technology

Article

Figure 2. COD and ammonia removal percentages in SBRs during pulse and continuous inputs. Graphs show the capabilities of the SBRs to remove nutrients and the upsets that occurred in each time period. 14 d, 6 d, and 2 d pulse spikes of Ag or AgNPs were followed by continuous Ag addition to investigate microbial resilience to Ag.

suggesting that their dispersion should have been similar.10 The nanoparticles also have similar total organic carbon concentrations, but GA AgNPs were nearly twice the size of the citrate AgNPs (Table S3, Supporting Information). GA AgNPs may have adsorbed to sludge biomass in the wastewater matrix more readily than the citrate coating. Although GA AgNPs are classified as hydrophobic, hydrophilic zones within the coating may be able to bind to sludge flocs.21 The hydrophobic and hydrophilic properties of the polymer coating may have bound to biomass more readily. Since effluent total Ag concentrations were so low, it is likely that Ag and AgNPs precipitated out of solution as they bound with wastewater ligands. Similarly, Hou et al.22 found that 90% of citrate AgNPs remained in SBRs and subsequent treatment cycles would thus be exposed AgNPs attached to sludge flocs until sludge wasting occurred. Kiser et al.23 showed that AgNPs were removed not only by aggregation and sedimentation but also by biosorption onto heterotrophic sludge biomass. Dissolved Ag concentrations were minimal in the SBRs suggesting that most of the Ag as AgNO3 or Ag+ released from the AgNPs may have been combined with the abundance of chloride or sulfide groups normally present in sludge biomass.24 Reactor Treatment Efficiency. The presence of AgNPs and AgNO3 in wastewater SBRs was found to significantly impact heterotrophic and autotrophic activity in the SBRs (p < 0.05). During the 0.2 ppm spikes, the removal efficiencies of COD and ammonia decreased rapidly immediately following pulse additions but stabilized within 3−5 d (Figure 2). It is interesting to note that the total Ag concentration decreased in all treatments by at least 40% in 24 h after the first pulse addition. This rapid reduction in total Ag may be linked to the rapid treatment recoveries observed at the 0.2 ppm exposures. During the 2 ppm spikes, the SBRs were not able to recover as quickly.

that observed during the pulse addition experiments, GA AgNPs had the lowest total Ag concentration in the SBRs compared to the other treatments, which may have been due to attachment with settling sludge flocs. Effluent total and dissolved Ag was also measured periodically to determine how much Ag was flowing out of the reactors. The effluent total Ag concentration was very low and not statistically different from the no Ag control (p < 0.05). Effluent dissolved Ag concentrations were below detection. After the 0.2 ppm spikes, the concentration of Ag was increased to 2 ppm for citrate AgNPs, GA AgNPs, and AgNO3. Immediately following the first 14 d spike (1 min after Ag addition), the concentrations of total Ag were measured as 1.12 ± 0.06, 0.73 ± 0.15, and 0.77 ± 0.26 ppm for citrate AgNPs, GA AgNPs, and Ag as AgNO3, respectively. After 24 h, the total Ag concentration had reduced by 99.8 ± 0.05% in the SBRs for all Ag reactors. Effluent total and dissolved Ag measurements were not statistically different from no Ag control SBRs (p < 0.05), indicating that Ag was either precipitating and aggregating at the bottom of the reactor or sorbing to sludge biomass. Sorption of Ag to biomass has been shown to be heavily prevalent in wastewater, with up to 98% removal of functionalized AgNPs from supernatant.19 The same trends in Ag concentrations were observed for the 2 and 6 d spikes. The continuous phase followed the pulses, and GA AgNPs were still consistently lower in concentration than citrate AgNPs and AgNO3. Dissolved Ag samples were not statistically different from no Ag controls (p < 0.05). Overall, total Ag concentration decreased most rapidly in SBRs receiving GA AgNPs in their influent. This result is surprising since GA AgNPs are typically known to be very stable due to their steric and electrostatic stabilized coating which keeps them well dispersed.20 In previous experiments, we found that GA and citrate AgNPs had similar zeta potentials, 973

dx.doi.org/10.1021/es403640j | Environ. Sci. Technol. 2014, 48, 970−976

Environmental Science & Technology

Article

COD removal was moderately affected by AgNPs, especially following the first 0.2 ppm pulse addition. Removal percentages dropped from 99% to 71%, 80%, and 90% for citrate AgNPs, GA AgNPs, and Ag as AgNO3, respectively, at the end of the first pulse addition treatment cycle. However, after 3 d, the COD removal recovered significantly and efficiencies were 92%, 95%, and 92% for citrate AgNPs, GA AgNPs, and Ag as AgNO3, respectively. The COD removal efficiencies were not statistically different from the control (p > 0.05). The same trend was observed for the second and third 0.2 ppm pulse additions, with treatment efficiencies of 77%, 81%, and 79% following the third 0.2 ppm pulse addition. Overall, Ag from AgNO3 did not affect COD removal significantly more than citrate or GA AgNPs. This trend may be due to a handful of reasons. Ag from AgNO3 may have bound quickly with chlorides and sulfides, reducing toxicity to heterotrophic bacteria that are mainly responsible for COD removal.25 It is also possible that the heterotrophic bacterial community was slowly adapting to the presence of Ag, similarly to their ability of developing antibiotic resistance.26 The continuous spike of 0.2 ppm Ag and AgNPs did cause a slight drop in treatment efficiency after the first cycle from 96% to 77%, 81%, and 79% for citrate AgNPs, GA AgNPs, and Ag as AgNO3, respectively; however, the reactors recovered rapidly and removed COD by 90% or more after 3 d. This trend was also observed with the addition of the 2 ppm spikes. After the first 14 d spike, no large decreases in COD efficiency were observed. This trend continued until the continuous spike, where removal efficiencies dropped slightly to 92%, 76%, and 82% for citrate AgNPs, GA AgNPs, and Ag as AgNO3, respectively, after 2 d but the SBRs recovered to steady state efficiency after 4 d. The improvement in recovery may have been linked to a more drastic shift in microbial community structure in response to the constant flow of AgNPs. Please refer to the “Effects of AgNPs on Microbial Community Structure” section for a more detailed discussion on community dynamics. Ammonia removal was more strongly affected by Ag as AgNO3 than AgNPs. In general, ammonia removal efficiencies dropped immediately after the pulse additions but were able to recover similarly to COD removal. After the first 0.2 ppm pulse addition, ammonia removal decreased from 98% down to 97%, 83%, and 32% for citrate AgNPs, GA AgNPs, and Ag as AgNO3, respectively. These results suggest that Ag as AgNO3 caused a statistically greater drop in ammonia removal than AgNPs (p < 0.05). This effect is congruent with our previous study that also found Ag as AgNO3 to be lethal to N. europaea, a model wastewater ammonium oxidizing bacterium, at 0.2 ppm.10 GA and citrate coated AgNPs were found to have similar mortality effects at concentrations of 2 ppm total Ag. Treatment recovery of the Ag as AgNO3 spiked reactor took longer than the AgNP treated reactors. However, after 8 d, all reactors had recovered and were removing in excess of 88% of the influent ammonia. The second and third pulse additions had virtually no effect on ammonia removal, indicating that the nitrifying microbial community in the SBRs likely adapted following the initial spike to maintain high levels of activity even in the presence of AgNPs. This observation is possibly due to the presence of functionally redundant bacteria in the SBRs. Additional information is presented later in the manuscript discussing changes in microbial community structure. The continuous spike of 0.2 ppm Ag did not significantly affect ammonia removal, with the exception of the Ag as AgNO3

treatment that lowered removal to 64% on d 55, but overall, the reactors were able to recover. The 2 ppm spike of Ag was more disruptive and sustaining than the previous lower concentration Ag addition. Similarly, Ag as AgNO3 had the most significant impact on ammonia removal and caused a decrease in removal from 95% to 64% in the Ag as AgNO3 after 2 d. The reactor recovered after 14 d but consistently underperformed with each additional spike and finally showed constant poor performance in the continuous phase, with an average ammonia removal of 83%. On the other hand, the other forms of Ag (i.e., AgNPs) did not appear to have a strong impact on ammonia removal and were not statistically different from the no Ag control during the 2 ppm spikes. Overall, these data suggest that AgNPs were less toxic to ammonia oxidizing bacteria than AgNO3. One explanation for why AgNPs could be less toxic is the effects due to their coating, which may have encouraged precipitation and limited dissolution of toxic Ag species. Nitrate and nitrite concentrations were consistently monitored but no strong inhibitory trends were apparent in those results. Please refer to Supporting Information for more details on nitrate/nitrite measurements (Figure S4). Effects of AgNPs on Microbial Community Structure. T-RFLP was performed on DNA samples from the SBRs and analyzed for clustering, diversity, and abundance. Microbial communities prior to the Ag spike were compared to communities after Ag was added. The universal 16S SSU rDNA T-RF chromatograms were compared, which provides a representation of bacterial communities in the SBRs. Principle coordinate ordinations (Bray−Curtis similarity) help visualize the clustering of different time points and treatments (Figure 3). The SBR communities before Ag spiking all clearly clustered but, as Ag additions became more frequent, the Ag treated communities became less similar from the control communities. The first noticeable trend in clustering occurred following the first pulse addition (14 d). Grouping of CA14 and GA14 and distancing of AG14 was observed. After the second and third 0.2 ppm pulse additions (6 and 2 d), the GA and CA communities shifted more toward the AG14 community. During the continuous phase 0.2 ppm Ag addition, the AgNP communities heavily grouped with AG14 and shifted from the first spike to a tight cluster. The Ag as AgNO3 exhibited the most immediate effects on microbial community, while the AgNPs showed a more gradual shift. The control communities did not group with the Ag treated communities, but there was some variation in their similarity to each other. This could be due to the variability of the controls, which is common in wastewater sludge. The diversity was relatively higher than other Ag treated communities for some time points such as day 32, shown by Simpson 1-D diversity indices. In general, the AgNP additions lowered diversity in most treatments during the first 32 d of Ag addition (pulses only). The effects of individual AgNPs is not immediately discernible from the ordination plots. However, diversity indices (Table S4, Supporting Information) suggest that GA AgNPs lowered diversity more than citrate AgNPs at several treatment time points. The citrate and GA AgNP treated SBRs never reached the same diversity they had prior to Ag addition, but they did appear to improve in diversity by the last day of the continuous 0.2 ppm Ag addition. Most notably, diversity dropped dramatically immediately following the first 14 d pulse addition of 2 ppm Ag. While Ag as AgNO3 decreased the diversity the most (i.e., 94.6%), GA AgNPs and citrate AgNPs 974

dx.doi.org/10.1021/es403640j | Environ. Sci. Technol. 2014, 48, 970−976

Environmental Science & Technology

Article

adaptation and Ag-ligand formation will reduce their overall impacts. While there may be brief upsets in treatment at 0.2 ppm Ag, it is likely that wastewater microbial communities have enough functional redundancy to recover quickly. If concentrations reach 2 ppm total Ag, major plant upsets could occur and biological nutrient removal could be disrupted. However, the likelihood of such a scenario is unknown currently since the concentration of AgNPs from consumer products is not well documented. These results suggest that adsorption, aggregation, and low dissolved Ag were important factors that dictated how AgNP could impact microbial community structure and function in wastewater sludge. In general, the microbial community diversities decreased with Ag spikes but were able to survive Ag addition more readily than nonacclimated microbial communities. Our novel approach to dosing and spiking SBRs with Ag and AgNPs demonstrates that nitrification can rebound rapidly, especially at the lower Ag concentrations. It should be noted that a microbial community’s ability to survive under high metal concentrations will be linked to the development of heavy metal resistance and, therefore, the effects of AgNPs on treatment efficiency could vary between different nanoparticles. Despite this, our results clearly indicate that wastewater microbial communities are able to recover from Ag and AgNP additions, depending on the concentration, but their dynamics may shift. Studies are ongoing to characterize more types of AgNPs, the development of microbial resistance to AgNPs, and the genera specific effects of AgNPs.



ASSOCIATED CONTENT

S Supporting Information *

Detailed information regarding experimental methods, nutrient and Ag measurements, Ag speciation, AgNP characterization, and microbial community diversity. This material is available free of charge via the Internet at http://pubs.acs.org.

Figure 3. Bray−Curtis principal coordinate ordination diagrams of (A) 0.2 ppm Ag additions and (B) 2 ppm Ag additions show clustering of microbial communities. Each sample is denoted by its treatment (AG = Ag as AgNO3, GA = gum arabic AgNPs, CA = citrate AgNPs, C = no Ag control) and the phase (0 = before Ag addition, 14 = 14 d pulse, 6 = 6 d pulse, 2 = 2 d pulse, cont = continuous Ag addition).



AUTHOR INFORMATION

Corresponding Author

*Phone: 919-660-5208; fax: 919-660-5219; e-mail: [email protected] duke.edu.

also caused significant decreases (92.6% and 32.4% reduction, respectively). Zheng et al.27 observed a similar decrease in SBRs treated with TiO2 nanoparticles. However, to our knowledge, the present study is the first demonstration of a significant decrease in microbial diversity in SBRs receiving AgNPs. Another study performed by Yang et al.28 investigated the affected microbial genera through pyro-sequencing. Their results concluded that nitrifying bacteria were particularly susceptible to upset by AgNPs and Ag as AgNO3, which is supported by the temporary loss of nitrification observed in this study. Bray−Curtis similarity indices (Table S5, Supporting Information) also confirmed that there was very little similarity between the microbial communities of the 0.2 ppm continuous Ag addition and the communities characterized during the 2 ppm 14 d pulse Ag addition. The 2 ppm pulse and continuous inputs of all Ag types drastically lowered diversity and caused major upsets in terms of microbial community. The similarity between the 0.2 ppm continuous addition communities and all of the first 2 ppm Ag spike communities approached zero, indicating a total shift in the bacterial community. Implications of AgNPs on Wastewater Treatment Performance. These data suggest that AgNPs may disrupt COD and ammonia removal initially, but eventually, microbial

Notes

This work has not been subjected to EPA review and no official endorsement should be inferred. The authors declare no competing financial interest.



ACKNOWLEDGMENTS This material is based upon work supported by the National Science Foundation (NSF) and the Environmental Protection Agency (EPA) under NSF Cooperative Agreement EF0830093, Center for the Environmental Implications of NanoTechnology (CEINT). Any opinions, findings, conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the NSF or the EPA. We thank our collaborators at CEINT for providing AgNPs and AgNP characterization to us.



REFERENCES

(1) (a) Chen, X.; Schluesener, H. J. Nanosilver: A nanoproduct in medical application. Toxicol. Lett. 2008, 176 (1), 1−12. (b) Fabrega, J.; Renshaw, J. C.; Lead, J. R. Interactions of silver nanoparticles with Pseudomonas putida biofilms. Environ. Sci. Technol. 2009, 43 (23), 9004−9009. 975

dx.doi.org/10.1021/es403640j | Environ. Sci. Technol. 2014, 48, 970−976

Environmental Science & Technology

Article

(2) Bystrzejewska-Piotrowska, G.; Golimowski, J.; Urban, P. L. Nanoparticles: Their potential toxicity, waste and environmental management. Waste Manage. 2009, 29 (9), 2587−2595. (3) Benn, T. M.; Westerhoff, P. Nanoparticle silver released into water from commercially available sock fabrics. Environ. Sci. Technol. 2008, 42 (11), 4133−4139. (4) Geranio, L.; Heuberger, M.; Nowack, B. The behavior of silver nanotextiles during washing. Environ. Sci. Technol. 2009, 43 (21), 8113−8118. (5) Farre, M.; Gajda-Schrantz, K.; Kantiani, L.; Barcelo, D. Ecotoxicity and analysis of nanomaterials in the aquatic environment. Anal. Bioanal. Chem. 2009, 393 (1), 81−95. (6) Gheju, M.; Pode, R.; Manea, F. Comparative heavy metal chemical extraction from anaerobically digested biosolids. Hydrometallurgy 2011, 108 (1−2), 115−121. (7) Bedard, C.; Knowles, R. Physiology, biochemistry, and specific inhibitors of CH4, NH4+, and co-oxidation by Methanotrophs and Nitrfiers. Microbiol. Rev. 1989, 53 (1), 68−84. (8) (a) Percival, S. L.; Bowler, P. G.; Russell, D. Bacterial resistance to silver in wound care. J. Hosp. Infect. 2005, 60 (1), 1−7. (b) Castellano, J. J.; Shafii, S. M.; Ko, F.; Donate, G.; Wright, T. E.; Mannari, R. J.; Payne, W. G.; Smith, D. J.; Robson, M. C. Comparative evaluation of silver-containing antimicrobial dressings and drugs. Int. Wound J. 2007, 4 (2), 114−122. (9) (a) Carlson, C.; Hussain, S. M.; Schrand, A. M.; Braydich-Stolle, L. K.; Hess, K. L.; Jones, R. L.; Schlager, J. J. Unique cellular interaction of silver nanoparticles: Size-dependent generation of reactive oxygen species. J. Phys. Chem. B 2008, 112 (43), 13608− 13619. (b) Morones, J.; Elechiguerra, J.; Camacho, A.; Holt, K.; Kouri, J.; Ramírez, J.; Yacaman, M. The bactericidal effect of silver nanoparticles. Nanotechnology 2005, 16 (10), 2346−2353. (10) Arnaout, C. L.; Gunsch, C. K. Impacts of silver nanoparticle coating on the nitrification potential of Nitrosomonas europaea. Environ. Sci. Technol. 2012, 46 (10), 5387−5395. (11) Lee, P. C.; Meisel, D. Adsorption and surface-enhanced Raman of dyes on silver and gold sols. J. Phys. Chem. 1982, 86 (17), 3391− 3395. (12) Yin, L.; Cheng, Y.; Espinasse, B.; Colman, B. P.; Auffan, M.; Wiesner, M.; Rose, J.; Liu, J.; Bernhardt, E. S. More than the ions: The effects of silver nanoparticles on Lolium multif lorum. Environ. Sci. Technol. 2011, 45 (6), 2360−2367. (13) Meyer, J. N.; Lord, C. A.; Yang, X. Y.; Turner, E. A.; Badireddy, A. R.; Marinakos, S. M.; Chilkoti, A.; Wiesner, M. R.; Auffan, M. Intracellular uptake and associated toxicity of silver nanoparticles in Caenorhabditis elegans. Aquat. Toxicol. 2010, 100 (2), 140−150. (14) (a) Wastewater Technology Fact Sheet: Sequencing Batch Reactors. EPA 832-F-99-073. Environmental Protection Agency: Washington, D.C., 1999. (b) Kos, P. Short SRT (solids retention time) nitrification process/flowsheet. Water Sci. Technol. 1998, 38 (1), 23−29. (15) Zeng, R. J.; Lemaire, R.; Yuan, Z.; Keller, J. Simultaneous nitrification, denitrification, and phosphorus removal in a lab-scale sequencing batch reactor. Biotechnol. Bioeng. 2003, 84 (2), 170−178. (16) Lukow, T.; Dunfield, P. F.; Liesack, W. Use of the T-RFLP technique to assess spatial and temporal changes in the bacterial community structure within an agricultural soil planted with transgenic and non-transgenic potato plants. FEMS Microbiol. Ecol. 2000, 32 (3), 241−247. (17) Culman, S. W.; Gauch, H. G.; Blackwood, C. B.; Thies, J. E. Analysis of T-RFLP data using analysis of variance and ordination methods: A comparative study. J. Microbiol. Methods 2008, 75 (1), 55− 63. (18) Culman, S.; Bukowski, R.; Gauch, H.; Cadillo-Quiroz, H.; Buckley, D. T-REX: Software for the processing and analysis of TRFLP data. BMC Bioinf. 2009, 10 (1), 171. (19) Kiser, M. A.; Ryu, H.; Jang, H.; Hristovski, K.; Westerhoff, P. Biosorption of nanoparticles to heterotrophic wastewater biomass. Water Res. 2010, 44 (14), 4105−4114. (20) (a) Balantrapu, K.; Goia, D. V. Silver nanoparticles for printable electronics and biological applications. J. Mater. Res. 2009, 24 (9),

2828−2836. (b) Lin, S.; Cheng, Y.; Liu, J.; Wiesner, M. R. Polymeric coatings on silver nanoparticles hinder autoaggregation but enhance attachment to uncoated surfaces. Langmuir 2012, 28 (9), 4178−4186. (21) Song, J. E.; Phenrat, T.; Marinakos, S.; Xiao, Y.; Liu, J.; Wiesner, M. R.; Tilton, R. D.; Lowry, G. V. Hydrophobic interactions increase attachment of gum arabic- and PVP-coated Ag nanoparticles to hydrophobic surfaces. Environ. Sci. Technol. 2011, 45 (14), 5988− 5995. (22) Hou, L. L.; Li, K. Y.; Ding, Y. Z.; Li, Y.; Chen, J.; Wu, X. L.; Li, X. Q. Removal of silver nanoparticles in simulated wastewater treatment processes and its impact on COD and NH4 reduction. Chemosphere 2012, 87 (3), 248−252. (23) Kiser, M. A.; Ryu, H.; Jang, H. Y.; Hristovski, K.; Westerhoff, P. Biosorption of nanoparticles to heterotrophic wastewater biomass. Water Res. 2010, 44 (14), 4105−4114. (24) Adams, N. W. H.; Kramer, J. R. Silver speciation in wastewater effluent, surface waters, and pore waters. Environ. Toxicol. Chem. 1999, 18 (12), 2667−2673. (25) (a) Ratte, H. T. Bioaccumulation and toxicity of silver compounds: A review. Environ. Toxicol. Chem. 1999, 18 (1), 89− 108. (b) Trevors, J. T. Silver resistance and accumulation in bacteria. Enzyme Microb. Technol. 1987, 9 (6), 331−333. (26) Silver, S. Bacterial silver resistance: Molecular biology and uses and misuses of silver compounds. FEMS Microbiol. Rev. 2003, 27 (2− 3), 341−353. (27) Zheng, X.; Chen, Y.; Wu, R. Long-term effects of titanium dioxide nanoparticles on nitrogen and phosphorus removal from wastewater and bacterial community shift in activated sludge. Environ. Sci. Technol. 2011, 45 (17), 7284−7290. (28) Yang, Y.; Quensen, J.; Mathieu, J.; Wang, Q.; Wang, J.; Li, M.; Tiedje, J. M.; Alvarez, P. J. J. Pyrosequencing reveals higher impact of silver nanoparticles than Ag+ on the microbial community structure of activated sludge. Water Res. 2014, 48 (0), 317−325.

976

dx.doi.org/10.1021/es403640j | Environ. Sci. Technol. 2014, 48, 970−976