Dissolution and transformation of silver nanoparticles in flow systems: Sulfidation and effect of flow rate

Dissolution and transformation of silver nanoparticles in flow systems: Sulfidation and effect of flow rate Lorenzo Sanjuan-Navarro a,b,* , Sergio Cort´ es-Bautista b , Melanie Vital a , Frank von der Kammer a a Department of Environmental Geosciences, Centre for Microbiology and Environmental System Science, University of Vienna, 1090, Vienna, Austria b MINTOTA Research Group, Department of Analytical Chemistry, Faculty of Chemistry, University of Valencia, 46100, Burjassot, Valencia, Spain HIGHLIGHTS GRAPHICAL ABSTRACT •Continuous-flow tests showed a direct correlation between flow rate and NPs dissolution. •Oxygen availability was found to be a critical factor. •Sulfidation occurred in two stages, forming amorphous AgxSy more soluble than Ag2S. •Natural organic matter and MOPS buffer significantly influenced dissolution and transformation pathways. •SEM and μ XRF analyses confirmed morphological changes and sulfidation products. ARTICLE INFO Keywords: Engineered nanoparticles Silver nanoparticles Continuous-flow methodology Dissolution Sulfidation ABSTRACT The release of engineered nanoparticles (ENPs) into the environment is an emerging concern with significant implications for organism exposure due to the possible toxicity caused by these materials. Understanding the dissolution and transformation of silver nanoparticles (AgNPs) under dynamic flow conditions is critical, as these processes directly govern the mobility, persistence, and bioavailability of this material in natural aquatic systems, ultimately influencing its ecological risk and environmental impact. In this work, a methodology based on continuousflow systems to assess the dissolution and transformation behavior of nanomaterials (NMs) under environmentally relevant conditions has been applied specifically to AgNPs. In this way, AgNPs nanopowder has been analyzed employing different environmental conditions (presence of oxygen, background electrolyte, types of buffers and NOM,). Results confirmed the expected dependency of AgNPs dissolution on the oxygen availability, corroborating previous batch assays. The evaluation of the AgNPs sulfidation in oxic and anoxic regimes were applied under different flow rates. AgNPs sulfidation is a complex process that takes place in two steps, oxidation followed by the AgS formation. Two different sulfide compounds * Corresponding author at: MINTOTA Research Group, Department of Analytical Chemistry, Faculty of Chemistry, University of Valencia, 46100, Burjassot, Valencia, Spain. E-mail address: [email protected] (L. Sanjuan-Navarro). Contents lists available at ScienceDirect Science of the Total Environment journal homepage: www.elsevier.com/locate/scitotenv https://doi.org/10.1016/j.scitotenv.2025.180939 Received 9 July 2025; Received in revised form 13 October 2025; Accepted 7 November 2025 Science of the Total Environment 1006 (2025) 180939 0048-9697/© 2025 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY license ( http://creativecommons.org/licenses/by/4.0/ ). with different solubilities can be formed: amorphous AgxSy which is related to an elevated Ag ion release, and high ordered Ag x S y phase which create a protective layer that reduces the ion release. Overall, the proposed continuous-flow methodology offers a reliable tool to investigate the dissolution and transformation of ENPS under environmentally relevant conditions. It provides valuable insights into the mechanisms that govern nanoparticles behavior, particularity during the initial stages of reaction. 1. Introduction In recent years, research efforts regarding the assessment of the environmental behavior of engineered nanoparticles (ENPs) have significantly increased (Bathi et al., 2021). The study of ENPs is mandatory due to their widespread applicability across various research fields, as well as their potential to reach different environmental compartments like surface waters, soils and sediments (Martinez et al., 2021). Moreover, their implications in these environments are important, considering their possible toxicity and bioavailability (Alves Jorge de Souza et al., 2019). Among ENPs, silver nanoparticles (AgNPs) are considered one of the most important types due to their unique physicochemical properties (Lekamge et al., 2018; Xu et al., 2020). The commercial production of Ag nanoproducts in textiles, personal care products, paints, antiseptic sprays, and medical commodities have raised concerns regarding environmental hazards and their impact on human health (McGillicuddy et al., 2017; Marin et al., 2020; Ong et al., 2022). Moreover, depending on their applications, the concentration of this nanomaterial (NM) in the environment can vary over orders of magnitude from ng•L −1 in surface waters to mg•L −1 in wastewater effluents, depending on the compartment and emission source (Kakakhel et al., 2021; Gottschalk et al., 2013). To better understand the toxicity and bioavailability of AgNPs, it is crucial to investigate their behavior in complex natural environment (Jahan et al., 2017). As introduced previously, ENPS, particularly AgNPs, can reach aquatic environments during their life cycle, through sewage discharge, atmospheric dry and wet deposition, and waste leachate (Ding et al., 2019; Fornest et al., 2020; Shevlin et al., 2018). The presence of AgNPs in these environments can lead to harmful effects on plant growth and development (Yokesh et al., 2014), toxicity to microbial systems in the soil (Schlich et al., 2013), and different degrees of damage to the liver, spleen, lungs and kidneys in animals due to the uptake (Wei et al., 2015). Moreover, in these environments, NMs can undergo changes in their properties, thereby altering their behavior. The most important processes affecting NMs fate in the environment are dissolution, chemical transformations and (hetero-)agglomeration (Alves Jorge de Souza et al., 2019). Different physical, biological and chemical parameters (Shevlin et al., 2018) play a crucial role in these processes, impacting their viability. For example, environmental conditions (e.g. dissolved oxygen, DOM, ionic strength, pH, sulfide, phosphate or chloride ions), presence of natural substances (bio and natural colloids), particle morphology and surface chemistry, as well as seasonal variations in water chemistry, strongly influence on the transformation processes of AgNPs (Kang et al., 2023; Dale et al., 2015; Liu et al., 2014). The determination and evaluation of these processes under realistic environmental conditions requires large liquid to solid ratios due to the low concentrations of NM in e.g. surface waters. These requirements can be met by specialized batch experiments or continuous-flow reactor experiments. The latter guarantees constant reactant concentration and no depletion. In batch studies, different methodologies have been applied with suitable results. Studies (Zhang et al., 2018; Sanjuan-Navarro et al., 2020) have corroborated the strong dependence of particle size on the dissolution rate, showing that, when normalized to surface area, smaller AgNPs release more Ag + ions than larger ones. The pH is another parameter that influences the AgNPs dissolution, where acidic and neutral conditions accelerated aggregation and dissolution, while alkaline conditions stabilized the particles (Fernando et al., 2019). Moreover, a new batch methodology was developed to study ENMs transformations under relevant environmental exposure scenarios, characterizing and evaluating AgNPs sulfidation processes using electron microscopy techniques (Stetten et al., 2024). The effect of AgNPs sulfidation on eco-toxicity was also studied demonstrating a decrease after the transformation process by sulfide interaction (Levard et al., 2013; Prathinkra Devi et al., 2015). While batch experiments have provided valuable insights into AgNPs dissolution and transformation, they are limited by static conditions that do not reflect the dynamic nature of natural aquatic environments. In real systems, continuous water exchange, chemical composition gradients, and fluctuating redox conditions strongly influence nanoparticle stability and speciation. Flow-through systems better capture these processes by simulating advective transport and sustained chemical fluxes, thereby offering a more environmentally realistic framework to assess the fate of AgNPs. Thus, shifting from batch to flow conditions is essential to evaluate the persistence, mobility, and ecological risks of AgNPs under scenarios that approximate natural waters (Kang et al., 2023). Furthermore, most studies employ a unique analytical technique to carry out the characterization and assessment of AgNPs transformations. Combining different techniques could provide a more comprehensive understanding of AgNPs transformation processes in real aquatic environments (Zhang et al., 2018; Prathinkra Devi et al., 2015). For this reason, in this work, AgNPs dissolution and transformation by sulfidation have been studied in continuous-flow reactors under aerobic and anaerobic conditions. Different flow rates have been employed to reproduce the possible environments where AgNPs can be released. The AgNPs transformation reactions have been characterized by ion release in the solution phase and with μ XRF and SEM on the solid phase. 2. Materials and methods 2.1. Materials and reagents Silver nanoparticles (AgNPs, Sigma 484,059-5G) with a particle diameter <150 nm were used to assess the solubility and dissolution of Ag nanopowder, based on environmental relevance and experimental considerations (dry powder). The same material was also employed to evaluate the sulfidation process. Additionally, sodium sulfate trihydrate (Na 2 S - 3H 2 O, VWR) and sodium nitrate (NaNO 3 , Merck) were used to prepare different media for the dissolution and transformation assays. The pH of the solutions was adjusted using sodium hydroxide (NaOH, Merck) and nitric acid (HNO 3 , VWR). Furthermore, humic acid (HA, Fluka) and 3-(N-morpholino)propanesulfonic acid (MOPs, 1.0 M, Thermo Fisher) were used as buffers to maintain a pH of 7 in the transformation experiments. Copper nitrate (Cu(NO 3 ) 2 , Merck) was used for the filter membrane pretreatment. The sulfide concentration was monitored using a colorimetric method, which involved N,N-dimethyl-p-phenylenediamine sulfate (Merck), ferric chloride (FeCl 3 ⋅6H 2 O, (Acros Organics)) and hydrochloric acid (HCl, VWR) as reagents. Nanoparticles retention was carried out using Hydrosart filter (cellulose, 10 KDa, 47 mm, Sartorius) as membrane. L. Sanjuan-Navarro et al. Science of the Total Environment 1006 (2025) 180939 2 2.2. Instruments The total concentration of dissolved AgNPs was quantified using an ICP-MS 7900 (Agilent Technologies) equipped with an ASX-500 series autosampler (Agilent Technologies). The ICP-MS was fitted with a quartz cyclonic spray chamber and a MicroMist nebulizer (Agilent Technologies). Throughout all ICP-MS analyses, the plasma power was set to 1550 W, the plasma gas flow was 15.00 L⋅min −1 , the nebulizer gas flow was 1.08 mL⋅min −1 , and the auxiliary gas flow was 0.90 mL⋅min −1 . For the different studies carried out, the total ion content was quantified against an external calibration curve prepared in 0.54 % nitric acid for silver, using solutions of 1000 mg⋅L −1 (Inorganic Ventures) as initial calibration standards, prepared in 5 % nitric acid. The internal standard was prepared from a 1000 mg⋅L −1 ICP-MS rhodium standard and diluted to a concentration of 10 μ g⋅L −1 with 0.54 % nitric acid. The internal standard was added to the samples online using a T-piece. Blanks and spiked samples were included in all analyses for quality control. Fig. 1. Experimental systems employed for (a) assessment of AgNPs dissolution using the NM as nanopowder and (b) assessment of AgNPs nanopowder transformations. The NM is sandwiched between membranes which keep the NM in position, reduce dead volume and prevent damage of the ultrafiltration membrane. After loading the filter assembly, the assembly is turned upside-down to allow for an upward flow regime. This removes air pockets in the filter assembly. L. Sanjuan-Navarro et al. Science of the Total Environment 1006 (2025) 180939 3 To reduce carry-over, a rinsing procedure with 3 % and 6 % nitric acid was performed after all samples. The limit of detection (LOD) of the measured ion in the sample was calculated as LOD =3⋅SD, where SD is the standard deviation of ten blank samples. The LOD value obtained was 3.5⋅10 −3 μ g⋅L −1 . The effect of sulfidation in AgNPs powder after complete drying was evaluated using an ICP-OES 5110 instrument (Agilent Technologies) coupled to an SPS 4 autosampler (Agilent Technologies) to quantify the total concentration of dissolved silver ions. The optimal operating conditions were obtained using a plasma gas flow of 12.00 L⋅min −1 , a nebulizer gas flow of 0.65 mL⋅min −1 , and an auxiliary gas flow of 1.20 mL⋅min −1 with a plasma powder of 1200 W. For this analysis, a calibration curve generated from standard solutions of 1000 mg⋅L −1 (Inorganic Ventures) in 0.54 % nitric acid was employed. Blanks and spiked samples were included in all analyses for quality control. The same rising process used in the ICP-MS was applied. In this case, the LOD was 6.2 μ g⋅L −1 . Sulfide concentration was monitored using a Lambda 35 UV–Vis spectrometer (PerkinElmer), following a colorimetric method explained in Section 2.4. The results were obtained using a wavelength of 670 nm, with the corresponding baseline correction. To monitor and characterize the NM composition, an X-ray analytical microscope ( μ XRF) (XGT-7000, Horiba) was used. The instrument was equipped with two software-controlled X-ray guide tubes with diameters ranging from 10 μ m to 1.2 mm. The samples were analyzed at atmospheric pressure. Scanning electron microscopy (SEM) was performed using a scanning electron microscope (FEI Quanta 3D FEG-SEM), equipped with a Schottky-type field-emission electron gun and an EDAX Pegasus Apex IV detector system, which includes an EDAX Digiview V EBSD camera for crystallographic orientation determination. During the analysis, the electron beam was set to an accelerating voltage of 15 kV, with a probe current of approximately 148 pA in analytical mode. Samples were prepared using carbon coating (the sputter coating) to prevent the accumulation of electrostatic charge. 2.3. Continuous-flow systems The systems used to evaluate the dissolution and transformation processes are illustrated in Fig. 1. Each system consisted of depots containing the media, a pump, a reactor with the NM, and a collection section. Fig. 1a shows the setup for dissolution assays, while Fig. 1b depicts the system for transformation studies. For dissolution experiments (Fig. 1a), the NM powder was weighed and deposited onto the surface of a 10 KDa molecular weight cellulose filter, which served as the primary retention membrane. To secure the NM, a cellulose nitrate filter with a pore size of 0.1 μ m was placed on top. In order to minimize potential mechanical damage caused by flow forces, an additional pair of glass microfiber filters was positioned on each side of the setup, providing structural support and protecting the integrity of the primary membranes. Then, the entire filter–NM assembly was inserted into the reactor in an inverted orientation (rotated 180◦). The pore size of the filters was selected according to the flow direction of the medium within the system, ensuring that the retention occurred at the level of the 10 KDa membrane. The medium was circulated through the system, and outlet samples were recollected at different time intervals. For transformation assays (Fig. 1b), the system was adapted to operate under either anaerobic or aerobic conditions. A dual arrangement was required to simultaneously maintain oxygen-free (anaerobic) conditions, preventing sulfide volatility, and oxygen (aerobic) conditions, enabling NM oxidation. Additionally, the system could be entirely operated under anaerobic conditions to evaluate the effect of oxygen on the transformation process. The mixture of both depots in each part determined the final sulfide concentration, which interacted with the AgNPs. Anaerobic conditions were achieved by purging the solution medium with argon during and after preparation. AgNPs powder was weighed and treated as described for dissolution assays, and water or sulfide solutions were flowed across the solid. Samples were recollected at different time intervals, and after each assay, sulfide concentration was monitored using a UV–Vis test, and silver dissolution was measured by ICP-MS. Additionally, the pH of the outlet (each collected tube) was measured. The filter surface, containing the treated AgNPs, was analyzed using an X-ray analytical microscope. 2.4. Procedures Prior to all assays, the filter membranes were pretreated with dissolved copper nitrate to reduce possible interaction of Ag + ions with the membrane surface and prevent specific losses through adsorption. The copper pretreatment involved using Cu(NO 3 ) 2 at concentration of 100 mg⋅L −1 and 1 mg⋅L −1 , as well as NaNO 3 at 1 mM, following the steps shown in Table 1. These solutions flowed through the filter membrane prior to the weighing of powder NM at 1 mL⋅min −1 . Fig. S.2.b shows the dissolution assays conducted at 1 mL•min −1 with and without copper nitrate pretreatment. In the assays with pretreatment, an initial increase in silver ion concentration at the outlet is observed during the early stages of the reaction. This increase is not seen in the assays without pretreatment, likely due to the adsorption of Ag + ions onto the membrane surface. These results confirm that copper nitrate pretreatment effectively blocks potential binding sites, preventing initial Ag + losses. To evaluate the effect of sulfidation in AgNPs powder after complete drying of solid, a dissolution assay was conducted in a static batch system. In this assay, AgNPs powder and sulfide-AgNPs powder obtained from sulfidation with Na 2 S at 5.0 mM were used. First, 10 mg of NM powder was weighed and added to Teflon bottles, followed by the addition of 1 mL of ethanol as a pre-wetting step (Jensen et al., 2011). The volume was completed to 100 mL with MilliQ water, adjusted to either pH 5 or pH 7 using MES buffer (10 mM). The ionic strength was regulated with 10 mM NaNO 3 . The batches were sonicated for 15 min. Samples were collected at time 0, 1, 3, 24 h, 48 h and 96 h. Time point 0 refers to the situation after mixing the NMs with the buffered MilliQ water, followed by 15 min of sonication and 15 min of centrifugation. All other time points correspond to the moment of sampling and 15 min of centrifugation. During the assay, the batches were agitated on an orbital shaker at 200 rpm. Collected samples were filtered using Pierce™ Protein Concentrator tubes (Thermo Fisher, Austria - PES membrane, 10 KDa (MWCO)). However, ions can be adsorbed on the surface of PES membranes, potentially modifying the ion concentration in the output solution (Gr¨ af et al., 2023). To prevent ion losses due to membrane adsorption, a pretreatment of the membranes was carried out. This involved introducing 2 mL of 100 ppm Cu 2+ solution into the tubes which saturate potential binding sites on the membrane, followed by immediate centrifugation at 4500g for 5 min. Subsequently, 5 mL of MilliQ water was added for rinsing, and the tubes were centrifuged again at 4500g for 5 min. Samples (4 mL) were then recollected. During sample processing, tubes were centrifugated at 4500g for 15 min. The bottom of the tubes was pre-loaded with 100 μ l of 6 M HNO 3 to acidify the samples after filtration and immediately stabilize the dissolved ions at a low pH. To monitor the sulfide concentration in the solution and in the collected samples, a colorimetric assay was used. A mixed diamine reagent, consisting of N,N-dimethyl-p-phenylenediamine sulfate and ferric chloride in the ratios specified in (Cline et al., 1969), was prepared in high-purity hydrochloric acid at a concentration of 50 %. For the assay, 1.5 mL of sulfide samples were mixed with 0.12 mL of reagent. After 20 min, the solutions were measured using a UV–Vis spectrometer with the appropriate water dilution. To ensure the utmost precision and reliability of the results, dissolution and transformation experiments were performed in duplicate. L. Sanjuan-Navarro et al. Science of the Total Environment 1006 (2025) 180939 4 3. Results and discussion 3.1. Silver nanoparticles dissolution. Effect of oxygen availability and flow rates Given the wide range of applications of AgNPs, evaluating their dissolution behavior is an initial step in understanding their potential transformations and changes in properties. For this reason, a continuous-flow test was developed to establish the dissolution of AgNPs using 1 mM NaNO 3 in ultra-pure water as the medium and various flow rates under different experimental conditions, aiming to accurately replicate potential dissolution effects . This medium was selected to provide a background electrolyte to maintain a constant ionic strength within the range typical of natural freshwaters (0.1–5 mM). Nitrate was chosen because it is environmentally relevant and does not strongly complex with silver or lead to precipitation, allowing the study to focus on the intrinsic dissolution behavior of AgNPs without interference from the ion added. Fig. 2 presents the results of the AgNPs dissolution experiments, employing flow rates of 5.0 mL⋅min −1 , 1.0 mL⋅min −1 , and 0.2 mL⋅min −1 , under both aerobic (Fig. S.2.a) and anaerobic conditions (Fig. S.2.b). This experiment was performed employing the system described in Fig. 1.a. The different flow rates tested resulted in varying concentrations of Ag + ions at the outlet. Under aerobic conditions, Ag + concentrations were initially high for all flow rates but decreased significantly over time. At a flow rate of 5.0 mL•min −1 , elevated Ag + concentrations were maintained throughout the experiment. In contrast, a flow rate of 1.0 Table 1 Copper pretreatment employed in the filter membranes to carry out the NMs dissolution and transformation assays using the powder NM. Step 1: Step 2: Step 3: Cu(NO 3 ) 2 100 mg⋅L −1 Cu(NO 3 ) 2 1 mg⋅L −1 NaNO 3 1 mM Flow (mL⋅min −1 ) NM as powder Previous weighing 10 min 10 min 15 min 1 Fig. 2. Dissolution assays of AgNPs nanopowder in high concentration obtained from continuous-flow experiments under different experimental conditions. (a) and (b) Aerobic assays using different flow rates: 5.0 mL⋅min −1 (blue), 1.0 mL⋅min −1 (grey) and 0.2 mL⋅min −1 (orange). In yellow assay without pretreatment conditions. (c) and (d) Anaerobic assays using different flow rates: 5.0 mL⋅min −1 (blue), 1.0 mL⋅min −1 (grey) and 0.2 mL⋅min −1 (orange). L. Sanjuan-Navarro et al. Science of the Total Environment 1006 (2025) 180939 5 mL•min −1 led to a more rapid and pronounced decrease in Ag + concentration per liter. Under anaerobic conditions, Ag + concentrations remained consistently low across all flow rates. Only at 0.2 mL•min −1 was a slightly higher Ag + concentration observed at the beginning of the experiment. The analysis of the initial Ag + release rates at 5.0 mL•min −1 and 1.0 mL•min −1 within the first 60 min reveled similar trends. To quantify this, the initial release profiles (up to 60 min, excluding the first recollected point) were fitted by linear regression. The calculated slopes were −0.039 ±0.009 μ g•L −1 •min −1 at 5.0 mL•min −1 and −0.0417 ±0.008 μ g•L −1 •min −1 at 1.0 mL•min −1 , showing no significant differences at 95 % confidence level. This indicates that the early-stage release is governed primarily by surface oxidation kinetics rather than transport limitations in the boundary layer. However, these results do not provide a truly comparable assessment, as the total volume of water treated varied with each flow rate. To address this, the data were normalized over time, and the silver release rate was expressed in terms of micrograms of Ag + per second ( μ g•s −1 ) (Fig. 2). For flow rates of 5.0 mL⋅min −1 , 1.0 mL⋅min −1 , and 0.2 mL⋅min −1 , the total dissolution percentages relative to the initial amount of treated AgNPs in aerobic conditions were (352 ±17)⋅10 −4 %, (26 ±5)⋅10 −4 % and (94 ±0.7)⋅10 −5 %, respectively. These results reveal a direct correlation between NM dissolution and flow rate, with higher dissolution observed at higher flow rates. This can be explained by the aqueous boundary layer (ABL) diffusion model, which describes the diffusion of a compound on the particle's exterior (Grathwohl et al., 1998). In this case, increasing the flow rate reduces the ABL thickness, facilitating faster diffusion of Ag + ions from the NM into the medium. The study by Henkel et al. (2023) highlights the impact of flow rate on the release kinetics of specific compound in NMs and micromaterials. These findings indicate that the dissolution process is mainly governed by mass transport phenomena, which are strongly enhanced at higher flow rates. Moreover, the effect of copper pretreatment conditions was also analyzed under aerobic conditions. Fig. 2.b shows the experiments performed with the pretreatment step (grey) and without it (yellow). In this case, a higher initial ion release was observed in the experiment with the cooper pretreatment, suggesting its effective role in minimizing interaction between the metal ions and the filter membrane surface. Under anaerobic conditions (Fig. 2.c and d), the total AgNPs dissolution was lower than under aerobic conditions. The total AgNPs dissolution percentages were (55 ±3)⋅10 −4 %, (3 ±2)⋅10 −4 %, and (66 ±3)⋅10 −5 % for flow rates of 5.0 mL⋅min −1 , 1.0 mL⋅min −1 , and 0.2 mL⋅min −1 , respectively. In this case, the negligible dissolution is related to the anoxic conditions, which prevent NPs oxidation, thereby inhibiting Ag + release. This highlights the key role that oxygen content plays in AgNPs oxidation, followed by dissolution. 3.2. Silver nanoparticles transformations. Effect of sodium sulfide During their life cycle, AgNPs interact with different environmental compounds that can alter their properties. One significant interaction is with sulfides, which leads to the sulfidation of AgNPs. This transformation produces diverse effects on the behavior of nanomaterials, necessitating thorough monitoring (Levard et al., 2011). To examine the different effects of sulfide, a flow-through sulfidation system for AgNPs under aerobic and anaerobic conditions, as described in Fig. 1.b, was developed. In this case, the experiment was performed without pretreatment conditions, using sodium sulfide at concentrations of 0.5 and 5.0 mM at pH 7. Based on the results obtained, during the first hours of the reaction, an increase in the Na 2 S concentration in the medium led to an increase in AgNPs dissolution, observed in both aerobic and anaerobic conditions (Fig. 3.a). This effect is surprising, as the formation of Ag 2 S should typically result in low solubility. However, this observation could be explained by the formation of amorphous Ag x S y on the NPs surface, Fig. 3. Silver ion concentration ( μ g⋅L −1 ) obtained from AgNPs dissolution as a function of time for each experiment carried out using H 2 O pH 7 (grey), Na 2 S 0.5 mM at pH 7 (blue) and Na 2 S 0.5 mM at pH 7 (orange) in aerobic and anaerobic conditions employing a flow rate of 1.0 mL•min −1 . Schematic representation of sulfidation process. L. Sanjuan-Navarro et al. Science of the Total Environment 1006 (2025) 180939 6 which has a higher apparent solubility compared to more highly ordered Ag x S y phases. This amorphous formation is supported by the SEM results, presented later in Section 3.3 (Fig. 7), providing direct experimental evidence in addition to the previously reported data (Gogos et al., 2017, 2018; Ma et al., 2014). Another possible explanation could be the short contact time between the medium and the NMs. Complexes of Ag + and S 2− may pass through the membrane before forming nuclei large enough to be retained, despite the oversaturation of Ag 2 S. Aerobic and anaerobic assays demonstrated that the presence of oxygen plays a key role in the formation of amorphous Ag x S y compound and, subsequently, in the release of silver. Under aerobic conditions, a significantly higher amount of amorphous compound was produced due to the enhanced oxidation effect on the AgNPs surface. Consequently, the formation of a larger amount of this compound also led to an increase in dissolution. In contrast, under anaerobic conditions, the formation of amorphous Ag X S y compound was minimal, and the release of Ag + ions dropped considerably, suggesting either suppression of the sulfidation process or a different reaction pathway. In previous batch experiments (Stetten et al., 2024), the formation of hollow silver sulfide spheres was observed, although the experimental time frame spanned days rather than hours. The combined findings suggest a rapid formation of a silver sulfide shell, while the subsequent reaction, in which dissolved silver migrates through the shell and is either precipitated or transported away as a sulfide complex, occurs at a comparatively slower rate. A schematic of the sulfidation pathway of AgNPs based on these results is presented in Fig. 3.b. In this model, the oxidation step enables the formation of silver ions, which interact with the sulfide present in the medium. This sulfide can form a protective layer around the NPs surface, potentially preventing further ion release. However, part of this protective layer consists of amorphous silver sulfide, which promotes the formation of small clusters and exhibits distinct dissolution kinetics that can influence NPs dissolution. To corroborate these findings, a dissolution test was conducted after complete drying in static batch system. In this experiment, standard AgNPs powder and sulfide-AgNPs powder obtained from sulfidation with 5.0 mM Na 2 S were dispersed in MilliQ water at pH 5 and pH 7. Samples were collected at time intervals of 0, 1, 3, 24 h, 48 h and 96 h in order to evaluate the total NP dissolution. As can be seen in Fig. 4, sulfidized AgNPs exhibited a slower dissolution of NPs due to the formation of a protective layer of Ag 2 S on the NP surface. At pH 7, the NM dissolution was lower than at pH 5, which is consistent with pre-test studies performed. In both cases, the effect of sulfidation was observed. However, the dissolution was not fully suppressed, indicating an excess of silver that did not react with sulfide, supporting the hypothesis of a core-shell structure. Using the pretreatment conditions described previously, a silver sulfidation assay was conducted using sodium sulfide at concentrations of 0.1 and 0.5 mM at pH 7 and different flow rates (0.2 mL⋅min −1 , 1.0 mL⋅min −1 and 5.0 mL⋅min −1 ) under aerobic and anaerobic conditions. The ionic strength was 1 mM of NaNO 3 . Figs. 5 and S.4 present the results related to these assays. The results are explained by considering the global assay and two time-sections that help understand the different effects involved in the transformation process. According to the results (Fig. S.3), time-section A (T.S.A) from 0 to 60 min shows a drastic decrease in dissolution behavior, whereas time-section B (T.S.B) from 60 to 240 min shows stable dissolution. Under aerobic conditions (Fig. 5), the sulfidation assay produced different results depending on the flow rate used. At a flow rate of 5.0 mL⋅min −1 , the protective layer effect was evident, resulting in a reduction of dissolution by more than 50 %. At a flow rate of 0.2 mL⋅min −1 , the dissolution of amorphous silver sulfide was the dominant process, leading to an increase in the amount of silver ion. Finally, at a flow rate of 1.0 mL⋅min −1 , the overall effect of sulfidation was negligible. Focusing on the flow rate of 1.0 mL⋅min −1 , the time-section graphs showed different results from the overall pattern described earlier. Both processes, as explained previously, were observed. In the 0–60 min section, the dissolution of amorphous silver sulfide predominated, while in the second section (up to 240 min), the protective layer effect became apparent, confirming the dissolution reduction associated with the sulfidation process. As shown in Fig. S.4, under anaerobic conditions, the oxidation/ sulfidation process was negligible, as indicated by the extremely low ion concentration in the outlet. The ion dissolution remained similar across all cases, whether using water or varying concentrations of sodium sulfide. In the aerobic experiments, pH measurements at the outlet at different times ranged from 5.8 to 6.9. For the anaerobic experiments, the pH at the outlet was similar, ranging from 5.5 to 6.8. This clearly demonstrates the need for a suitable buffering agent in the media to maintain a stable pH. However, a buffer could potentially influence the transformation reactions in various ways, such as by complexing with released ions, surface adsorption, or even reactive interaction. For this reason, the effects of humic acid (HA) and 3-(N-morpholino) propanesulfonic acid (MOPS) buffer on the dissolution and transformation processes associated with sulfidation were evaluated. Figs. 6 and S.5 show the results of AgNPs dissolution employing HA and MOPS as buffers at different flow rates and oxygen conditions, alongside the pH Fig. 4. Batch dissolution experiment at pH 5 (a) and pH 7 (b) (MES buffers) using AgNPs and sulfide-AgNPs. Experiments were carried out over a period of 96 h. The initial concentration of NM in the batches is 100,000 μ g⋅L −1 . L. Sanjuan-Navarro et al. Science of the Total Environment 1006 (2025) 180939 7 values at the outlet. The results revealed a behavior consistent with previous observations. The assay at 5.0 mL•min −1 exhibit higher dissolution per liter compared to the assay at flow rate of 1.0 mL•min −1 , which showed the lowest dissolution per liter using both buffers. However, a significant increase in the concentration of dissolved Ag + (% dissolution) was noted compared to prior assays that did not employ buffers. This effect is attributed to an increased oxidation rate of AgNPs when higher concentrations of HA or MOPS are present in the system. This phenomenon can be attributed to the effective adsorption of HA onto the surface of NPs, inducing both steric and electrostatic stabilization, which facilitates their dispersion. The abundant functional groups in HA play a crucial role in the transformation of NPs. As a result, HA interacts efficiently with AgNPs, enhancing their dissolution (Stankus et al., 2011). As the concentration of HA increases, more functional groups become available to form complexes with AgNPs (or with Ag 2 O on the surface of AgNPs) (Daniel et al., 2019). These robust complexes may weaken the surface Ag – Ag and Ag – O bonds, thereby accelerating the dissolution of AgNPs through processes promoted by organic ligands (Liu, 2021). The effect of MOPS exhibits a parallel behavior, suggesting a potential complexation of metal ions with MOPS organic molecules (Table 2). The sulfidation process (Figs. 6.b and S.5.b) yielded results consistent with previous observations. Notably, a greater reduction in dissolution was observed at a flow rate of 5.0 mL⋅min −1 . Conversely, at a flow rate of 0.2 mL⋅min −1 , the presence of amorphous silver sulfide became evident, leading to increased dissolution compared to the assay without sulfide when HA was used as buffer. Employing a flow rate of 1.0 mL⋅min −1 , the effect of the protective layer was noticeable but less pronounced. This is attributed to higher dissolution in the initial stages of the assay, influenced by the presence of the amorphous compound. The pH measured at the outlet (Figs. 6 and S.5) correlates with the dissolution effect. Higher pH values were observed at higher dissolution rates, which is related to the quantitative consumption of protons during the oxidation step. When using HA as a buffer, the pH values obtained were higher than those when employing MOPS. These differences could be attributed to the minor buffering capacity of HA in the system and the potential secondary transformations of the nanomaterial related to the multiple interactions of NOM with the metal (Khort et al., 2022). In the case of MOPS, the pH remained close to 7, indicating a strong buffering effect and showing a gradual decrease that correlated with the dissolution values. In sulfidation assays using HA as a buffer (Fig. 6), pH values were slightly higher at the initial sampling times compared to previous assays, likely due to increased dissolution associated with the amorphous sulfide compound. In contrast, the use of MOPs as buffer (Fig. 6.4) provided more stable and less basic pH at the beginning, indicating lower dissolution. Under anaerobic conditions, the observed pH values were lower than those under aerobic conditions, likely due to the negligible oxidation step and, consequently, the insignificant consumption of protons. 3.3. Analysis by XRF and SEM To complete the study of AgNPs transformations under aerobic and anaerobic conditions, the NMs were analyzed using μ XRF and SEM to determine and characterize the sulfidation process (Figs. S.6 and 7). For this analysis, AgNPs processed with 1 mM NaNO 3 solution without sulfide served as reference blanks. AgNPs treated with 0.5 mM Na 2 S in the presence of 1 mM NaNO 3 were used as the experimental samples. In both cases, the transformation assay was performed at a flow rate of 5.0 mL•min −1 . For the μ XRF analysis, the filter surface was examined directly using a glass support at atmospheric pressure. In the case of SEM, the NM was Fig. 5. Percentage of silver ion dissolved from AgNPs in the sulfidation process employing different concentration of sodium sulfide (0.1 and 0.5 mM) and different flow rates (0.2 mL⋅min −1 , 1.0 mL⋅min −1 and 5.0 mL⋅min −1 ) in aerobic conditions. (a) Total range of time, (b) Time-section from 0 to 60 min and (c) Time-section from 60 to 240 min. L. Sanjuan-Navarro et al. Science of the Total Environment 1006 (2025) 180939 8 collected from the filter surface and mounted on a pin using an adhesive, conductive carbon tab. As shown by the μ XRF microscope, AgNPs without sulfide treatment in aerobic conditions (Fig. S.6.b) only displayed bands corresponding to the present of silver. However, in the case of AgNPs treated with 0.5 mM Na 2 S in aerobic conditions (Fig. S.6.c), both silver and sulfur bands were detected, confirming that the sulfidation process had occurred. Fig. 7 presents the SEM results, showing the surface morphology of the NMs. In the case of AgNPs powder (Fig. S.7) and AgNPs without sulfide treatment (Fig. 7.a), the NMs exhibited various sizes and spheroidal or oval shapes. Similarly, the samples treated in anaerobic conditions with 0.5 mM Na 2 S (Fig. 7.b) displayed comparable morphologies, indicating that the NPs experienced negligible transformations under these conditions. However, when 0.5 mM Na 2 S (Fig. 7.c and d) was used as the medium under aerobic conditions, the morphology of the NM changed. In this scenario, particles with asymmetrical shapes and irregular profiles were observed. Additionally, the size distribution varied, resulting in smaller-sized particles and aggregates with a bulging appearance. The SEM images revealed surfaces that could be associated with crystalline (Fig. 7.c) and amorphous (Fig. 7.d) Ag x S y compounds. The observed morphologies closely resemble those reported in the literature for Ag 2 S NPs formed through both direct and/or indirect sulfidation of this NM (Zhang et al., 2018). EDAX analysis was conducted to verify the formation of the Ag X S y compound. For the NM powder and AgNPs without sulfide treatment (Figs. S.7 and 7.a), only bands corresponding to silver were observed. In contrast, samples treated with Na 2 S under aerobic conditions displayed spectra consistent with the formation of Ag 2 S (Fig. 7.c and d). The intensity ratio between the Ag L α and S K α emission lines (i.e., the height of the EDAX peaks) ranging between 1.7 and 2.2, suggesting nearcomplete sulfidation. Only a few areas of these samples did not show any transformation or exhibited lower ratios of the Ag x S y compound. Using the same procedure, samples treated with Na 2 S under anaerobic conditions (Fig. 7.b) were analyzed, and only silver bands were detected, with the sulfur band being negligible. These results are consistent with the previous assays, which indicate that the presence of oxygen is a key factor in the oxidation-transformation process. 4. Conclusions In this study, a comprehensive evaluation was conducted on the dissolution and sulfidation transformations of AgNPs within flowthrough systems, under varying environmental conditions such as oxygen presence, buffer type or NOM. The experiments were performed using optimal filters and pretreatment conditions. Furthermore, three distinct flow rates were applied: 0.2 mL⋅min −1 , 1.0 mL⋅min −1 , and 5.0 mL⋅min −1 , to evaluate the effect of flow rate on the dissolution and transformation processes. This methodology proves to be versatile for monitoring the Fig. 6. Dissolution/transformation assays developed employing HA as buffer a) without sulfide content in the medium, and b) with sulfide content in the medium. Different flow rates (0.2, 1.0 and 5.0 mL⋅min −1 ) and distinct oxygen condition (AE: aerobic conditions and AN: anaerobic conditions) are employed. The initial pH of medium is set at pH 7. In the right part, pHs measured in the outlet. L. Sanjuan-Navarro et al. Science of the Total Environment 1006 (2025) 180939 9