Introduction

In recent years, metal oxide nanoparticles, including nanosized Fe2O3, ZnO, Al2O3, TiO2, MgO, and CeO2, had been developed and applied extensively in a number of different areas including electronics, cosmetics, energy technologies, and biomedical, catalytic, and material applications, as well as environmental remediation [1,2,3]. The use of metal oxide nanoparticles had seen exponential growth in consumer and industrial products [4,5,6]. Among the major types of manufactured nanoparticles, aluminum oxide (Al2O3), one of the most abundantly manufactured metal oxides on a nanoscale, had been widely used in high-performance ceramics, cosmetic fillers, packing materials, polishing materials, semiconductor materials, paints, composite materials, and resins [7,8,9]. The global production of Al2O3 has been increasing year after year, and it is more economically favorable to use [10,11,12].

The extensive use of Al2O3 nanoparticles in consumer and industrial products has led to concerns about their potential environmental impacts during recent years [13, 14]. Firstly, if Al2O3 nanoparticles were released into water systems, these particles could potentially cause harm to aquatic organisms including algae species and aquatic invertebrates [15]. Moreover, there was concern over the possible adverse effects of Al2O3 nanoparticles on human health. Many existing researches indicated that Al2O3 nanoparticles exhibited adverse effects on cell proliferation and cell viability in human lung epithelial cells and A549 carcinoma cells through the biological and food cycle [16]. Understanding the aggregation and colloidal stability of manufactured Al2O3 nanoparticles in natural waters is essential to accurately evaluate their toxicity on aquatic organisms and human health [17,18,19]. Thus, investigations on the stability, migration, and aggregation propensity of Al2O3 nanoparticles in aquatic environments are of great significance.

To a certain degree, the particle diameter range of nanoparticles is an important parameter that determines their transport behavior, toxicity, and ultimate environmental fate in the water environment [20, 21]. According to the report, nanoparticle stability in aqueous environments depends on many parameters, including the environmental conditions (solution pH, ionic strength, and presence of other chemical compounds), nanoparticle concentration, and the nanoparticle intrinsic and surficial characteristics [22,23,24]. The unintended impact in aquatic environments, chemical reactivity, environmental fate, and behavior of nanoparticles and their possible interaction with environmental elements is thus a key issue in aquatic environmental risk assessment [25]. Therefore, this novel study investigates the aggregation and disaggregation behaviors of Al2O3 nanoparticles in different experimental conditions. In this paper, solution pH, electrolyte cations in different valences (Na+ and Ca2+), and humic acid (HA) were identified as major abiotic factors influencing the stability, aggregation, and disaggregation behaviors of Al2O3 nanoparticles in a complex aquatic system. In addition, dynamic light scattering (DLS) analysis was performed to explore the aggregation behaviors of Al2O3 nanoparticles and fluorescence spectra analysis was used to analyze the aggregation mechanisms. Information obtained in this study will be of fundamental significance for the understanding of the ultimate fate of inorganic colloids themselves and contaminants in complex environmental conditions [26].

Materials and methods

Materials

NaCl, CaCl2, NaOH, HNO3, and Al2O3 nanoparticles were obtained from commercial supplies (Shanghai Aladdin Biochemical Technology Co., Ltd.) and used without further purification. According to the manufacturer’s data, the physical characteristics of Al2O3 nanoparticles are as follows: purity > 99%, diameter = 50 nm, and crystalline phase of approximately 30%γ versus 70%θ. Commercial HA with purity > 90% was purchased from Sinopharm Chemical Reagent Co., Ltd. and used with prior separation and purification according to our previous work [27]; C (50.15%), H (4.4%), O (39.56%), S (0.49%), and N (5.38%) were the main components in the sample. Analytical reagent grade chemicals (NaCl, CaCl2, NaOH, and HNO3) were used along with ultrahigh-purity water (18.2 MΩ cm−1), which was purified with a Milli-Q equipment and used throughout the experiments to prepare solutions. Al2O3 colloidal particles were diluted to the desired concentrations with deionized Milli-Q water prior to use as a research object to study the aggregation and disaggregation behaviors of inorganic colloids [28]. Stock solutions of NaCl and CaCl2 were used as electrolytes to control the ionic strength of the solution and to study the effect of cation type on the stability of Al2O3 nanoparticles. An amount of 50 mg of purified HA was dissolved in 1L of Milli-Q water. The solution was then filtered over a 0.01-μm filter to remove all non-dissolved aggregates. The pH values of the solutions were adjusted by the addition of 0.01 M NaOH and/or HNO3 to obtain the desired pH conditions.

Zeta potential and hydrodynamic diameter measurement

The zeta potential and hydrodynamic diameter measurements of aggregating Al2O3 nanoparticles were performed with a Zetasizer nanoseries (Malvern Nano-ZS, Malvern, UK) instrument. In the experiment, He–Ne laser was used as a coherent source of monochromatic radiation at 633 nm and the scattering angle was 173° for all measurements [29]. For the aggregation experiment, the achieved solution with a certain concentration range of HA and Al2O3 nanoparticles was immediately placed in a disposable polystyrene cuvette to track the average diameter variation trend of the particles. Similar experimental procedure was also conducted to obtain the aggregation behaviors of Al2O3 nanoparticles in the presence of different metal cations (Na+ and Ca2+). During the initial aggregation process, each autocorrelation function was accumulated for 15 s, and the whole process lasted 4500 s. All the experiments were performed in triplicates and average values were presented in the work.

Fluorescence measurements

The fluorescence excitation-emission matrix (EEM) spectra and emission intensity changes of HA were measured on a HITACHI F-7000 spectrofluorometer with a 450W Xenon lamp. The emission wavelengths were set from 200 to 900 nm, while the excitation wavelengths were set from 200 to 600 nm with scanning number 25 and increment 10 nm [30]. For all the above fluorescence measurements, the excitation and emission slit widths were set at 5.0 nm with a scan speed of 1200 nm/min.

Results and discussion

Influence of pH on surface properties and stability of Al2O3 nanoparticles and HA

The surface charge modifications and stability of Al2O3 nanoparticles caused by adjusting the solution pH were determined using pH titration curves at 0.001 M NaNO3 ionic strength. As shown in Fig. 1A, there was an excellent relationship between the stability of Al2O3 nanoparticles and their electrophoretic properties. Meanwhile, the regularity of Al2O3 particles with different particle sizes being obtained by changing the pH value of the suspension can really reflect the dispersion and deposition of the aqueous suspensions of Al2O3 nanoparticles. As can be seen from Fig. 1A, the zeta potential values of Al2O3 nanoparticles were positive (30 ± 3.0 mV) and almost remained constant from pH 2.0 to 6.0, whereas the particles stayed the same size (220 ± 3 nm) through the entire pH range, which was larger than the original particle size (50 nm) of commercial Al2O3 particles, indicating an obvious aggregation of the colloidal particles in water solution [27]. Then, as the solution pH increased, the zeta potential value decreased to the point of zero charge (pHPZC) at a pH about 9.4 ± 0.1, whereas the particle size increased rapidly and the highest value of the z-average diameter obtained for the particle was 1700 nm at the pHPZC, which indicated a very easy and efficient aggregation of the particles when the positively charged surface was neutralized at 6.0 < pH < 9.4 [31]. By further increasing the pH, the magnitude of the zeta potential decreased as the pH increased and reached a minimum value of about − 36 mV at pH of approximately 12.0. Over the pH range 9.4 < pH < 12.0, one could also find that the particle size started to decrease rapidly and finally reached 220 ± 3 nm at pH 12.0 (it was similar to the initial particle size at pH 2.0 to 6.0), which was due to the amphoteric nature of Al2O3 nanoparticles [32]. Al2O3 was easier to dissolve gradually at high pH, and there was a relationship between the dissolution behavior and the decrease of the particle size.

Fig. 1
figure 1

Zeta potential variation and z-average diameter changes of Al2O3 particles (A) and HA molecules (B) as a function of pH. \(C_{\mathrm{Al}_2\mathrm{O}_3}\) = 100 mg/L, CHA = 50 mg/L

Full size image

The variations of the zeta potential of HA (50 mg/L) under different pH conditions were determined in a stock solution of NaNO3 to produce a background electrolyte concentration of 0.001 mol/L (Fig. 1B). Holding ionic strength constant, the zeta potential of HA decreased approximately linearly with increasing pH from − 14.0 mV at pH 2.0 to − 47.3 mV at pH 12.0, which was due to the existence of substantial amounts of carboxylic acid and phenolic functional groups on its surface [33]. To get an insight into a general trend of the aggregation properties of HA, the dependence of z-average on pH for HA at constant ionic strength (0.001 M NaNO3) is also illustrated in Fig. 1B. The z-average diameter of HA molecules was measured to be ~ 236 nm under acidic conditions; however, with the increase of solution pH, a slightly larger aggregate size (~ 293 nm) could be observed at pH > 8, which was because HA exhibited more than one type of macromolecular structure in aqueous solutions under different pH conditions [34]. Under acidic conditions, the molecules of HA showed globular and ring-like structures with a smaller aggregate size; however, the molecules became linear and/or sheet-like structures as the pH increases [34, 35]. Therefore, a slightly larger aggregate size could be observed under alkaline conditions as presented in Fig. 1B. Experimental data obtained in this work were shown to be in good agreement with experimental values obtained in other works [35, 36].

Effect of HA on the stabilization of Al2O3 nanoparticles under different pH conditions

pH < \(\mathrm{pH}_{PZC, \,Al_2O_3}\) (pH = 6.0)

The surface properties and particle size distribution of Al2O3 nanoparticle aggregates in the presence of HA were determined at pH = 6.0 (pH < pHPZC) to explore the capacities of HA in influencing the aggregation or stabilization of the particles. As shown in Fig. 2, before adding HA to the suspensions (CHA = 0 mg/L), Al2O3 nanoparticles showed a highly positive zeta potential value of 30.0 mV and a z-average diameter of 293 nm. However, after the addition of HA, the concentration influence of HA on zeta potential and the z-average diameter distribution curves were classified into three critical regions. In the first region (blue area in Fig. 2, CHA = 0–2.5 mg/L), the addition of HA led to a rapid increase in z-average diameter of Al2O3 nanoparticle aggregates. The zeta potential decreased with increasing HA concentration and reached the pHPZC at the concentration of 2.5 mg/L. Additionally, the z-average diameter reached a maximum value of 2157 nm, which indicated that the charge neutralization mechanism was responsible for the formation of the large aggregates [37]. With the increase of HA concentration as presented in the second region (yellow area in Fig. 2, CHA = 2.5–5.0 mg/L), it was established that charge inversion led to a significant decrease of the zeta potential and z-average diameter values until a HA mass concentration of 5.0 mg/L [38]. Finally, in the third region (red area in Fig. 2, CHA = 5.0–10.0 mg/L), the nanoparticles were in a diameter of 239 nm, and had a negative zeta potential of − 30.5 mV, enough to keep the nanoparticles stable in aqueous solutions without aggregating. The system was close to stabilization again due to the addition of higher HA concentrations.

Fig. 2
figure 2

The variations of zeta potential and z-average diameter of Al2O3 nanoparticles due to the presence of different concentrations of HA at pH 6.0. \(C_{\mathrm{Al}_2\mathrm{O}_3}\) = 100 mg/L

Full size image

pH = \(\mathrm{pH}_{PZC, \,Al_2O_3}\) (pH = 9.4)

The zeta potential and z-average diameter values of HA were found to be − 41.0 ± 2.0 mV and 290 ± 5 nm at pH = 9.4, respectively (Fig. 1B). However, as shown in Fig. 3, after the addition of HA, the system remained at equilibrium for a substantial amount of time (4500 s) between each measurement. We observed a rapid decrease in zeta potential of Al2O3 nanoparticles until a plateau at around − 32 mV when HA concentration was increased to 3 mg/L or greater during the negative surface charging process. Also, note that in this process, the fragmentation of the Al2O3 nanoparticles became significant, and the z-average diameter decreased and maintained at ~ 400 nm when HA concentration was in excess of 3 mg/L.

Fig. 3
figure 3

The variations of zeta potential and z-average diameter of Al2O3 nanoparticles due to the presence of different concentrations of HA at pH 9.4. \(C_{\mathrm{Al}_2\mathrm{O}_3}\) = 100 mg/L

Full size image

The kinetics of disaggregation of Al2O3 nanoparticles at pH = 9.4 was also measured as a function of time and HA initial concentration. The extent of the nanoparticle disaggregation was characterized by the variation of the zeta potential and z-average diameter values in all experiments. As shown in Fig. 4, the kinetic measurements revealed that the zeta potential and z-average diameter of Al2O3 nanoparticles decreased with the increase of HA concentration until one steady count at last. Typical rate curves with two different regimes for the aggregation of Al2O3 nanoparticles could be observed; the rate of disaggregation was comparatively high only during the initial period and decreased considerably later. Additionally, as can be seen, increasing HA initial concentration caused a larger amount of Al2O3 nanoparticles to disaggregate, and the maximum disaggregation was obtained at ≥ 7 mg/L HA with a final z-average diameter at 625 nm, which was a little larger than the initial Al2O3 particle size (293 nm), indicating a partial disaggregation of the nanoparticles [39]. The rapid decrease of zeta potential values of Al2O3 nanoparticles after the addition of HA indicated that the adsorption of negatively charged HA on Al2O3 nanoparticle surface resulted in a high electrostatic repulsion between Al2O3 nanoparticles, thus leading to a disaggregation of the nanoparticles [40].

Fig. 4
figure 4

A Time-dependent variation of the zeta potential of Al2O3 nanoparticles under different HA concentrations. B The extent of disaggregation of Al2O3 nanoparticles as a function of time at several HA concentrations. pH = pHpzc. \(C_{\mathrm{Al}_2\mathrm{O}_3}\) = 100 mg/L

Full size image

pH > \(\mathrm{pH}_{PZC, \,Al_2O_3}\) (pH = 11.0)

As shown in Fig. 5, the addition of HA caused a significant decrease in the zeta potential values at pH 11.0. However, there was a slight increase in the average size of the Al2O3 nanoparticles as HA concentration increased. Such experimental results showed that the adsorption of HA on Al2O3 nanoparticles was limited, and the effect of non-electrostatic van der Walls interactions played a significant role between HA and Al2O3 nanoparticles at pH > pHPZC, Al2O3 (pH = 11.0) [41].

Fig. 5
figure 5

The variations of zeta potential and z-average diameter of Al2O3 nanoparticles due to the presence of different concentrations of HA at pH 11.0. \(C_{\mathrm{Al}_2\mathrm{O}_3}\) = 100 mg/L

Full size image

Electrolyte-dependent aggregation behavior of Al2O3 nanoparticles

Aggregation kinetics and attachment efficiency studies

To examine the potential effects of electrolyte species on the aggregation of Al2O3 nanoparticles in the presence of HA, the aggregation kinetics and z-average diameter changes of Al2O3 nanoparticles were quantified in aqueous suspensions containing either monovalent (Na+) or divalent cations (Ca2+) as a background electrolyte through a combination of batch and column experiments. In the presence of Na+, an increase of the electrolyte concentration to 10 or 12 mM led to a rapid aggregation of the Al2O3 nanoparticles (Fig. 6A). However, further increase of the Na+ concentration over 12 mM had a weak effect on the enhancement of aggregation rate. But for the divalent electrolyte, a much lower Ca2+ concentration could induce a rapid aggregation of the particles; Ca2+ was far more effective in inducing Al2O3 nanoparticle aggregation. For instance, in the presence of 0.2 mM Ca2+, Al2O3 nanoparticles could quickly aggregate and finally reached a z-average size of ~ 1000 nm as presented in Fig. 6B, while the same aggregate size of the particles due to the presence of Na+ could be obtained at a much higher concentration of 10 mM as shown in Fig. 6A. As expected, divalent cations could make the surface charge of Al2O3 nanoparticles less negative due to the higher charge screening effect, which was beneficial for aggregation and thus inducing aggregation at a much lower concentration as compared to monovalent ions [41,42,43,44].

Fig. 6
figure 6

Aggregation profiles (A, B) and attachment efficiency studies (C) of Al2O3 nanoparticles as a function of Na+ (A) and Ca2+ (B) electrolyte concentration. CHA = 6 mg/L, pH = 6.0

Full size image

The Al2O3 nanoparticle attachment efficiency was determined from the aggregation kinetic profiles for Na+ and Ca2+ electrolyte species (Fig. 6C). For Na+ electrolyte, the attachment efficiency increased with the increase of electrolyte concentration until the Na+ concentration reached the critical coagulation concentration (CCC), which represented the minimum concentration of electrolyte that was needed to induce the occurrence of the aggregation [45]. In the presence of Na+, CCC value determined here for Al2O3 nanoparticles was about 40 mM. As the concentration of Na+ continued to increase, the negative surface charge of Al2O3 nanoparticles weakened significantly in the reaction-limited regime, thereby decreasing the electrostatic repulsion between particles and enhancing the attachment efficiency values [46]. When the Na+ concentration exceeded 40 mM, an increase in Na+ concentration had no effect on attachment efficiency. Under these conditions, the electrostatic repulsion between Al2O3 nanoparticles was completely suppressed and every collision between nanoparticles resulted in attachment [47]. In the presence of Ca2+, the curve for Al2O3 nanoparticle attachment efficiency was quite similar to the regime due to the addition of Na+, except that the CCC for divalent Ca2+ (0.5 mM) was much lower than that for Na+ (40 mM). Experimental results of CCC values obtained in this work have stated that multivalent ions were more effective in reducing the energy barrier between Al2O3 nanoparticles, and therefore promoting a more aggressive colloidal aggregation than monovalent ions.

Fluorescence spectra analysis

With the kinetic experiments indicating differences in aggregation mechanisms of Al2O3 nanoparticles in the presence of HA and different electrolyte species, comparing the fluorescence EEM spectra changes of HA due to the addition of electronic ions of Na+ and Ca2+ in Al2O3 suspensions can provide more insights into the respective aggregation mechanisms.

In order to obtain the optimum fluorescence spectra analytical conditions, the sorption experiment of HA on Al2O3 nanoparticles as a function of pH and HA concentration was investigated firstly in Fig. 7. It was observed that the sorption of HA on Al2O3 was highly dependent on solution pH and favored at lower pH values due to the electrostatic attractive interactions. In the case of CHA = 5 mg/L, pH < 9.0, the sorption percentage can be up to 99.5%, which indicated that HA was almost all adsorbed on the surface of Al2O3 nanoparticles. Therefore, according to the results of the sorption experiment, the optimum fluorescence spectra analytical conditions (pH = 7.0, CNa+ = 40 mM, CCa2+ = 0.5 mM, CAl2O3 = 100 mg/L, CHA = 5 mg/L) were determined as presented in Fig. 8.

Fig. 7
figure 7

The influence of pH on the sorption of HA with varying concentrations on Al2O3 nanoparticles. \(C_{\mathrm{Al}_2\mathrm{O}_3}\) = 100 mg/L, I = 0.01 mol/L NaNO3

Full size image
Fig. 8
figure 8

Fluorescence EEM spectra of HA (A) and HA aggregates in the presence of Na+ (B) or Ca2+ (C) metal ion at pH 7.0. CNa+ = 40 mM, CCa.2+ = 0.5 mM, \(C_{\mathrm{Al}_2\mathrm{O}_3}\) = 100 mg/L, CHA = 5 mg/L. D Changes in the intensity of the emission spectra of peak 1 and peak 2

Full size image

The EEM spectra of HA (Fig. 8A) and HA aggregates with the presence of electronic ions of Na+ and Ca2+ are shown in Fig. 8B and C, respectively. In Fig. 8A, two obvious fluorescent components were distinguished, including one carboxylic-like fluorophore (peak 1, excitation/emission = 250–275/425–475 nm), and one phenolic-like fluorophore (peak 2, excitation/emission = 300–325/425–450 nm) [48]. Additionally, the emission intensity maxima corresponding to peak 1 and 2 are listed in Fig. 8D as a comparison. Similar EEM contour plot of HA aggregates after the addition of Na+ ions in the suspension could be obtained in Fig. 8B; only the emission intensities of peaks 1 and 2 suffered a slight decrease as presented in Fig. 8D. However, after the addition of Ca2+ in the suspension (Fig. 8C), peak 2 disappeared and the emission intensity of peak 1 was weakened from 116 to 53 a.u. as shown in Fig. 8D, which indicated that the complex could remarkably quench the emission intensity of HA, and carboxyl and phenolic hydroxyl groups were both very effective in binding Ca2+ ions [49]. The results indicated that the Al2O3 nanoparticles may aggregate by more than one aggregation mechanism due to the presence of metal ions in different valence states. In the presence of monovalent Na+, increasing the ionic strength effectively suppressed the electrostatic double-layer repulsion; there was no repulsive energy barrier when two particles approached each other, and thus inducing the aggregation of Al2O3 nanoparticles. As for the divalent Ca2+, Ca2+ interacted with HA, whereas HA could be adsorbed on the surface of Al2O3 nanoparticles, and HA acted as a bridge between Ca2+ ions and Al2O3 nanoparticles. Therefore, the z-average diameter of Al2O3 nanoparticles could be increased through the function of complexing bridging between HA and Ca2+, which in turn resulted in the enhancement of aggregation [50]. There is a good coherence between the potential mechanism and the fact that the mapping of calcium was well associated with the aggregate positions of Al2O3 nanoparticles.

Conclusions

The data presented here indicated that combined influencing factors such as solution pH, HA, and ionic strength had great impacts on the stability of Al2O3 nanoparticles in the water environment system. An efficient aggregation of the particles can be observed at 6.0 < pH < 9.4 when the positively charged surface was neutralized. The largest aggregate size obtained at pH = pHPZC indicated there were close connections between the surface charge properties and stability of Al2O3 nanoparticles. The existence of the environmental aquatic concentration ranges of HA and electrolyte cations had great influence on the particle size, size distribution, and colloidal stability of Al2O3 nanoparticles. The complex behaviors of isolated or aggregated Al2O3 nanoparticles were dependent on the surface charge modification by adding different concentrations of HA in the suspension. From the theoretical point of view, the electrostatic interactions between the negatively charged surfaces made Al2O3 nanoparticles dispersed stably in the suspension at pH > pHPZC.. At pH < pHPZC, the electrostatic adsorption of the negatively charged HA macromolecules onto the surface of positively charged nanoparticles induced charge neutralization, where the maximum aggregation was achieved. However, a relatively strong charge inversion could be observed after a further addition of HA in the suspension, which indicated disaggregation became significant due to electrostatic repulsions between Al2O3 nanoparticles. Aggregation kinetic measurements and fluorescence EEM spectra analysis indicated that divalent Ca2+ can make the surface charge of Al2O3 nanoparticles less negative due to the higher charge screening effect, and induce a rapid aggregation of the particles at a much lower concentration as compared to monovalent Na+. The function of complexing bridging between HA and Ca2+ also resulted in the enhancement of aggregation. The Al2O3 nanoparticle stability was mainly controlled by the mechanism involving attractive and repulsive electrostatic interactions, van der Walls, steric effects, and the bridging effect of HA with electrolyte cations and nanoparticles. The study highlighted the influence of solution pH, HA, and electrolytes on the stability and colloidal aggregation process of metal oxide nanoparticles. These findings are useful in understanding the size change of the metal oxide nanoparticles, as well as the removal of the particles and HA from aqueous media, which is of great importance in assessing the environmental pollution of nanoparticles.