Amino-terminated SiO2–Al2O3 composite aerogels from fly ash for improved removal of Cu2+ and Pb2+ ions in wastewater: one-pot synthesis, excellent adsorption capacity and mechanism

In this study, by using a sol–gel grafting-atmospheric drying method, amino-terminated SiO2–Al2O3 composite aerogels, namely 3-aminopropyltriethoxysilane (APTES) or 3-(2-amino-ethoxy) propylmethyldimethoxysilane (AEAPMDS) modified SiO2–Al2O3 aerogels (AMSAAs), were synthesized from the fly ash and characterized by field-emission scanning electron microscopy, X-ray diffraction and X-ray photoelectron spectroscopy etc.. And the AMSAAs were verified as excellent adsorbents for removing heavy metal ions (Cu2+ and Pb2+ ions) from wastewater. The effects of modification conditions and testing parameters including pH value, adsorbent dose, initial ions concentration, adsorption time and temperature were systematically investigated. Results demonstrated that 0.2 mol/L APTES modified aerogels (0.2APTES-SAAs) possessed the best adsorption properties. Under the optimal pH value of 4.0–6.0 and the adsorbent dose of 0.4–0.6 g/L, the equilibrium adsorption capacities of Cu2+ and Pb2+ ions were as high as 195 mg/g and 500 mg/g within 20–30 min, respectively. The adsorption processes were agreed fairly well with Freundlich isotherm adsorption model and the pseudo-second-order kinetic model, which indicated that the adsorption processes were heterogeneous multilayer adsorption and controlled by the chemical reaction between AMSAAs and heavy metal ions. The obtained adsorption thermodynamic parameters (ΔH°, ΔS° and ΔG°) revealed that the adsorption processes were exothermic and spontaneous with decreased randomness at the solid–liquid interface. The excellent recyclability of as-prepared AMSAAs proved as economically promising adsorbents for practical applications.


Introduction
Heavy metals in wastewater are a common environmental problem with the development of economies and industries. Numerous heavy metals such as copper, lead, chromium, mercury, nickel, cadmium, etc. are toxic and can accumulate in human body through food chain and seriously affect human health (Ali et al. 2019;Arslan et al. 2020). Among the treatment methods of heavy metal ions in wastewater, adsorption method has attracted much more attention due to its convenience, high efficiency, economy and good selectivity (Chakraborty et al. 2020;Crini et al. 2019;. In general, the adsorption process is the result of both physical adsorption and chemical adsorption, and the former is related to its microstructure including specific surface area and pore structure etc., while the latter depends on its surface chemical properties. Therefore, the microstructure and surface chemistry of the adsorbent are the main factors determining the adsorption properties (Crini et al. 2019). Based on this, some ordered porous materials such as mesoporous carbon (Kazmierczak et al. 2021) and mesoporous silica (Maleki 2016) with large specific surface area and suitable surface valence bond characteristics stand out.
SiO 2 aerogel has a mesoporous three-dimensional network nanostructure with high specific surface area (> 500 m 2 /g), high porosity (> 80%), good thermal and chemical stability; especially, its surface groups can be regulated by functional group modification (Maleki 2016) for capturing heavy metal ions. SiO 2 aerogel has been traditionally synthesized from chemical reagents, and its surface is mostly Si-OH or Si-CH 3 groups. The high preparation cost and poor adsorption capacity for heavy metal ions become the main barriers to its practical application. Therefore, in recent years, many efforts have been taken on reducing the preparation cost by cheap raw materials such as rice husk ash (Feng et al. 2018;Chen et al. 2021), steel slag (Guzel Kaya et al. 2020) and other materials Ahmad and Mirza 2018;Jyoti et al. 2021), and functionalizing the aerogel surface by amino and/or sulfhydryl groups. Notably, functionalized SiO 2 aerogel via in situ polymerization Duraes 2017, 2019) or grafting Sertsing et al. 2018;Mirzaee et al. 2019) with amino groups is the main way to achieve high efficiency and chemical selective adsorption of heavy metal ions by strong coordination effects. Mirzaee et al. (2019) used water glass and (3-mercaptopropyl) trimethoxysilane (MPTMS) as raw materials to prepare sulfhydryl-modified SiO 2 aerogels by grafting method; the obtained aerogels presented certain adsorption capacities for Mn 2+ (1.38 mmol/L) and Zn 2+ (1.33 mmol/L). Sertsing et al. (Sertsing et al. 2018) found that the asprepared amino-modified SiO 2 aerogels from TEOS and APTES had a specific surface area of up to 1401.3 m 2 /g and good adsorption capacities for Ni 2+ (40.32 mg/g) and Cr 3+ (46.08 mg/g). However, most amino-modified SiO 2 aerogels are obtained by chemical reagents. How to use cheaper silicon source materials to synthesize amino modified absorbent with excellent adsorption properties in one step is still a challenging issue.
Fly ash is one of the most discharged solid wastes in the world, and its global average utilization rate is only 16%. A large amount of abandoned fly ash not only occupies cultivated land, but also seriously pollutes the environment. Till now, none of the work has been reported to prepare SiO 2 aerogel from fly ash and modified with amine groups directly to remove heavy metals. Given that the main components of fly ash are silicon oxide (40-50 wt%) and alumina (40-50 wt%), in this paper, amino-modified SiO 2 -Al 2 O 3 aerogels (AMSAAs) were successfully synthesized by a sol-gel grafting-atmospheric drying method using fly ash as raw material, 3-aminopropyltriethoxysilane (APTES) and 3-(2-amino-ethoxy) propylmethyldimethoxysilane (AEAP-MDS) as modifiers. The effects of experimental factors on the microstructure and the adsorption properties of as-prepared AMSAAs for Cu 2+ and Pb 2+ ions were discussed in detail. The adsorption isotherm, kinetics, thermodynamic and recyclability were also studied.

Synthesis of SiO 2 -Al 2 O 3 gel from the fly ash
The synthesis of SiO 2 -Al 2 O 3 gel is based on our previous work (Shen et al. 2020). The same amount of fly ash and NaOH were mixed evenly in a graphite crucible and then calcined in a muffle furnace at 773 K for 1 h. After cooling, 5 g calcined product was dissolved with 40 mL deionized water and 30 mL 5 mol/L HCl. After filtration, the pH value of the leaching solution was adjusted to 2.80 by 0.25 mol/L NH 3 ·H 2 O dropwise and the above solution aged for 1 day. After aging, the wet gel was obtained and washed with deionized water and ethanol for 3 times. Then, the gel was inserted into n-hexane and stirred for 12 h at 298 K and repeated 4 times. Finally, the SiO 2 -Al 2 O 3 gel was obtained after filtration for further use.

Modification of SiO 2 -Al 2 O 3 gel by amine
The SiO 2 -Al 2 O 3 gel was put into different concentrations of APTES/AEAPMDS modifier solution and stirred for 24 h at 298 K and repeated 2 times. After that, the gel was immersed into n-hexane and stirred for 6 h at 298 K and repeated 3 times. The final sample was obtained after filtration and dried at 318 K for 3 h, 338 K for 12 h and 348 K for 12 h. And they were labeled separately as AMSAAs (0.2 mol/L APTES or AEAPMDS modified SiO 2 -Al 2 O 3 aerogels (0.2APTES/ AEAPMDS-SAAs), 0.5 mol/L APTES or AEAPMDS modified SiO 2 -Al 2 O 3 aerogels (0.5APTES/AEAPMDS-SAAs) and 0.8 mol/L APTES or AEAPMDS modified SiO 2 -Al 2 O 3 aerogels (0.8APTES/AEAPMDS-SAAs)).

Characterization and analysis
The chemical composition, morphology and crystal phase of the samples were separately characterized by the X-ray fluorescence (XRF, XRF-1800, Shimadzu, Japan), field-emission scanning electron microscope (FE-SEM, S4800, Hitachi, Japan) and X-ray diffraction (XRD, D/ max-2500, Kratos, Japan). The chemical bond and functional group of as-prepared aerogels were identified with Fourier transform infrared spectroscopy (FTIR, FT-08, LUMEX). The valence bond structure was analyzed by X-ray photoelectron spectroscopy (XPS, K-Alpha, Thermofisher, USA). N 2 adsorption-desorption isotherms were tested; Brunauer-Emmett-Teller analysis (BET, Quadrasorb SI, USA) and BJH methods were used to calculate the specific surface area and pore size distribution of the samples. The pH measurements were done using pH meter (PHSJ-4F, China), and the zeta potential was measured by zeta potential analyzer (Zetasizer Nano Z, Malvern, UK). The concentrations of Cu 2+ or Pb 2+ ions were measured by an inductively coupled plasma-optical emission spectrometer (ICP-OES, OPTIMA 7000DV, Platinum Elmer, USA).

Batch adsorption experiment
In order to clarify the effects of type and concentration of modifiers, two kinds of above-prepared modified aerogels were used as adsorbents to study their adsorption properties for Cu 2+ ions, so as to determine the best modified adsorbent. The specific processes were as follows: 100 mg adsorbent was added to the sealed beaker containing 100 mL 50 mg/L Cu 2+ ion solution and then fully reacted under magnetic stirring. Part of the solution was taken out periodically. After centrifugation, the supernatant was taken to measure the concentration of Cu 2+ ions by ICP-OES, and the removal efficiency (R, %) of Cu 2+ by adsorbent was calculated according to formula (1).
In order to clarify the effects of adsorption parameters on removal of Cu 2+ and Pb 2+ ions, taking the above selected optimum adsorbent as the object, the effects of adsorption conditions such as the amount of adsorbent (0-1.2 g/L), pH value of the solution (2.0-8.0), reaction time (10-180 min) and initial ion concentration (10-500 mg/L) on the adsorption performance of Cu 2+ or Pb 2+ ions were further studied. Specifically, a certain amount of adsorbent was added to the sealed beaker containing 100 mL Cu 2+ or Pb 2+ ion solution with different concentrations and then fully reacted under magnetic stirring. Part of the solution was taken out periodically. After centrifugation, the supernatant was taken to measure the concentration of Cu 2+ or Pb 2+ ions by ICP-OES, and the removal efficiency (R) of Cu 2+ or Pb 2+ ions by the adsorbent was calculated according to formula (1). And the equilibrium adsorption capacity Q e (mg/g) was calculated as depicted in formula (2). (1) where, C 0 (mg/L) and C t (mg/L) are the concentration of Cu 2+ or Pb 2+ ions solution at the initial time and reaction time t, respectively. C e (mg/L) is the equilibrium concentration of Cu 2+ or Pb 2+ ion solution after adsorption, V (L) is the volume of solution and m (g) is the quality of adsorbents.

Zeta potential test
First, 0.1 g of the adsorbent was diluted with 0.1 L of deionized water. The solution was then dispersed for 2 h, and the pH of the solution was adjusted with 0.1 mol/L NaOH and HNO 3 solution with the pH ranges between 2.0 and 8.0 for test. The zeta potential values were plotted against pH value. The point of intersection of the resulting curve with the abscissa at which zeta potential value = 0 gave the point of zero charge value (pH ZPC ).

Desorption and recyclability of 0.2APTES-SAAs
One gram of 0.2APTES-SAAs was used as an adsorbent to treat 50 mL 50 mg/L of Cu 2+ ion solution for 20 min. After adsorption, the absorbed product was obtained by filtration and drying. And the 0.2APTES-SAAs was desorbed in 0.2 mol/L EDTA-2Na solution and stirred for 3 h. The above steps were repeated for 3 times, and 4 consecutive adsorption-desorption cycles were conducted.

Effects of type and concentration of modifiers on the microstructures
In order to investigate the effects of amino modifiers on the crystal structure and chemical bond characteristic, XRD patterns and FTIR spectra of the aerogels modified by different concentrations of APTES and AEAPMDS were conducted as shown in Fig. 1. All of the as-prepared samples were amorphous regardless of type and concentration of the amino modifiers, indicative of microstructure characteristics of aerogels. This phenomenon was also confirmed by the corresponding SEM images of the products as given in Fig. 2, in which similar three-dimensional network structures composed of nanoparticles with uniform size were also achieved. Moreover, as illustrated in Fig. 1b, the peaks around 1070 cm −1 were due to the vibrations of Si-O-Si (Shen et al. 2020). And the peaks at 800 cm −1 were the evidence of Al-O bonds (Shen et al. 2020;Ling et al. 2018). And the peaks at around 1562 cm −1 (not shown in the unmodified SAAs) were the evidence of -NH 2 bonds (Zhang et al. 2018). Similarly, in Fig. 1d the peaks at around 1070 cm −1 and 800 cm −1 were due to Si-O-Si and Al-O bonds (Shen et al. 2020;Ling et al. 2018), respectively. And the peaks at around 1572 cm −1 (not shown in the unmodified SAAs) were due to the bend mode of N-H bonds from AEAP-MDS (Zhang et al. 2018). The appearance of these peaks at 1562 cm −1 and 1572 cm −1 indicates that APTES and AEAP-MDS were successfully grafted into the framework of the SiO 2 -Al 2 O 3 composite aerogels. Generally, adsorbents with large specific surface area and suitable pore structure are favorable for the adsorption of metal ions in wastewater. So, the nitrogen  Table 1, respectively. All the as-prepared AMSAAs presented IV adsorption-desorption isotherms, indicative of mesoporous structures (Cychosz et al. 2017). In addition, with increasing the concentrations of APTES from 0.2 to 0.8 mol/L, the specific surface area of samples decreased from 355.44 to 211.48 m 2 /g, while the average pore size increased from 13.23 to 24.29 nm, and the pore volume increased from 1.05 to 1.25 cm 3 /g in Table 1. Almost the same trend appeared in the modification with AEAPMDS. However, by comparing the samples by different modifiers at the same concentration, it is found that the AEAPMDS had a greater impact on the specific surface area, while the APTES had a greater influence on the pore size of the as-prepared samples. The main reason may be ascribed to the different molecular structures of the two modifiers.

Effects of modification conditions on adsorption properties
A series of samples modified by APTES and AEAP-MDS were used as adsorbents to study the adsorption properties for Cu 2+ ions, and the experiment results are presented in Fig. 4. Both APTES and AEAPMDS could improve the adsorption properties of aerogels for Cu 2+ ions, and APTES modified aerogels were better, which might be due to the longer branched chain structure and larger steric hindrance during adsorption of AEAPMDS as depicted in Fig. S3. In addition, as can be seen from Fig. 4a, the removal efficiencies of Cu 2+ ions by different concentrations of APTES modified aerogels were basically the same; therefore, the 0.2 mol/L APTES modified SiO 2 -Al 2 O 3 aerogels (0.2APTES-SAAs) were selected as the object for the following experiments.

Effect of solution pH value
Researches revealed that copper and lead mainly exist in the form of Cu 2+ and Pb 2+ ions with the solution pH value less than 7.0. When the pH value is greater than 7.0, complex ions or precipitate may gradually generate (Huang et al. 2017). Therefore, the pH range of the solution in this section is controlled at 2.0-6.0. Figure 5a gives the change of removal efficiencies of Cu 2+ and Pb 2+ ions with pH value. As for Cu 2+ ions, with the solution pH increasing from 2.0 to 3.0, the removal efficiency increased rapidly from only 10 to 70%, then slightly enlarged to 90% when pH was up to 6.0. The similar trend also occurred in Pb 2+ ion adsorption, but quite remarkably, when the pH value increased from 2.0 to 3.0, the removal rate increased abruptly from 13 to 98%, then kept essentially unchanged through to 6.0. Based on the above analysis, the optimal solution pH value was controlled at 4.0-6.0. In order to analyze the effect of solution pH on the removal efficiencies of Cu 2+ and Pb 2+ ions, the zeta potential of 0.2APTES-SAAs in aqueous solution at different pH values was tested. Figure 5b shows the zero point charge value (pH ZPC ) of 0.2APTES-SAAs to be 8.0 which was greater than the pH value of solution. Therefore, the adsorbent would be positively charged. When the pH of the solution was at 3.0-5.0, the surface of 0.2APTES-SAAs was positively charged with a potential of up to 25 mV. As the pH value increased from 5.0 to 7.0, the surface potential decreased rapidly. The phenomena may be mainly ascribed to the following reasons: under the strong acidic conditions, the solution was rich in free H + ions, and -NH 2 groups of 0.2APTES-SAAs easily combined with H + ions to form -NH 3+ , which reduced the amount of -NH 2 interacting with Cu 2+ or Pb 2+ ions, then decrease the removal efficiency. Moreover, the electrostatic repulsion between -NH 3+ and Cu 2+ /Pb 2+ ions was one of the reasons for the low removal efficiency for Cu 2+ or Pb 2+ ions (Xu et al. 2020). However, when pH value increased to 6.0, the number of protonated amino groups decreased, and the electrostatic repulsion force reduced, so the removal efficiency of Cu 2+ or Pb 2+ ions increased accordingly.

Effect of adsorbent dose
The results for removal efficiencies of Cu 2+ and Pb 2+ ions with regard to adsorbent dose are shown in Fig. 6. The removal efficiencies for Cu 2+ and Pb 2+ ions were enhanced with increasing adsorbent dose from 0 to 20 mg (0 to 0.4 g/L). No surprising enhancement was observed for higher addictive amounts. Obviously, with enhancing adsorbent dose, more accessible adsorption sites were available which increased removal efficiency. And when the adsorbent dose was added more than 20 mg (0.4 g/L), the number of accessible adsorption sites that interacted with heavy metal ions gradually attained saturation, so the removal efficiency reached the maximum and stabilized at the maximum.  Figure 7a gives the plot of adsorption capacity as a function of initial ion concentration. It can be seen that with the initial concentration increasing from 50 to 850 mg/L, the adsorption capacity of Cu 2+ ions gradually increased to 190 mg/g (500 mg/L), then remained basically unchanged at about 195 mg/g (> 500 mg/L). Similarly, when the initial concentration of Pb 2+ ions was controlled from 50 to 900 mg/L, the adsorption capacity of Pb 2+ ions increased rapidly to 498 mg/g (700 mg/L) and then stabilized at about 500 mg/g (> 700 mg/L). Based on the above analysis, the equilibrium adsorption capacities of Cu 2+ and Pb 2+ ions were about 195 mg/g and 500 mg/g, respectively.

Effect of initial ion concentration and the adsorption isotherms
It should be noted that the adsorption capacity for Pb 2+ ions by 0.2APTES-SAAs was always much larger than that of Cu 2+ ions. The reason may be that the electron configuration of Cu atom is 3d 10 4s 1 , while that of Pb atom is 6s 2 6p 2 . Since Pb 2+ ions can provide more vacant orbitals than Cu 2+ ions, it is easier to form coordination bonds with N in the -NH 2 groups, that is, the affinity between -NH 2 and Pb 2+ ions was greater than that of Cu 2+ ions, so the adsorption capacity of Pb 2+ ions was larger (Xu et al. 2020).
In order to further study the thermodynamic mechanism of adsorption processes of Cu 2+ and Pb 2+ ions onto 0.2APTES-SAAs, Langmuir, Freundlich and Temkin adsorption isothermal models were used to establish the adsorption isothermal equilibrium (Hang et al. 2019;Hu et al. 2019;Sun et al. 2019;Oskui et al. 2019a). The ideal assumption of Langmuir model is that the surface of solid adsorbent is uniform, and the adsorption on the surface of adsorbent is monolayer adsorption; there is no interaction between the adsorbed molecules. Freundlich model believes that multilayer adsorption occurs, the surface of solid adsorbent is not uniform and there are certain interactions between adsorption molecules. Temkin model assumes that adsorption heat of all molecules in surface layer reduces linearly with the cover due to the interactions between absorbate and absorbent. And the Langmuir, Freundlich and Temkin isothermal models are given by Eqs.
where, C e (mg/L), Q e (mg/g), Q m (mg/g), b (L/mg) are the equilibrium concentrations of Cu 2+ or Pb 2+ ions, the adsorption capacity at equilibrium, the theoretical maximum adsorption capacity of adsorbents and Langmuir constant, respectively. K (mg/g) and n are the Freundlich constants. K T (L/mg) and B are Temkin constants.
(3) The adsorption isotherms and the calculated parameters from isotherms models are presented in Fig. 7b-d and Table 2. Compared with Langmuir and Temkin adsorption isotherm model, Freundlich model had better fitting results with the correlation coefficient as high as 0.992 for Cu 2+ ions and 0.978 for Pb 2+ ions. Accordingly, the Freundlich adsorption isotherm model was more suitable to describe the adsorption processes of 0.2APTES-SAAs for Cu 2+ and Pb 2+ ions. That is, both the adsorption processes of Cu 2+ and Pb 2+ ions onto the surface of AMSAAs were inhomogeneous multilayer adsorption.

Effect of adsorption time and the adsorption kinetic model
The effect of adsorption time on Cu 2+ and Pb 2+ ions removal by 0.2APTES-SAAs is shown in Fig. 8a. The removal efficiencies of Cu 2+ and Pb 2+ ions were both above 95% within 10 min and reached the maximum at 20 min, nearly kept stable at 100%. The reasons for the above result were closely related to the amino groups and large pore size structure of the absorbents. The strong coordination between the amino groups on the surface of the aerogels and metal ions in the solution provided a strong driving force for rapid adsorption. Meanwhile, the large average pore size also made it easier for Cu 2+ and Pb 2+ ions to diffuse into the adsorbents.
Generally, the adsorption process for Cu 2+ and Pb 2+ ions by AMSAAs can be divided into two steps: firstly, the diffusion of Cu 2+ and Pb 2+ ions to adsorbent surface, and secondly, the chemical reaction between Cu 2+ and Pb 2+ ions and amino groups on adsorbent surface, which corresponded to the pseudo-first-order (PFO) and pseudo-second-order (PSO) kinetic models, respectively. In order to better understand the adsorption process of metal ions by AMSAAs and determine whether the adsorption process is controlled by diffusion or chemical reaction, the PFO and PSO kinetic models were separately employed to fit the experimental data. The applied equations (Sorkhabi et al. 2021;Aghdasinia and Asiabi 2018) (6-7) are expressed in the following forms: where, Q e (mg/g), Q t (mg/g), k 1 (1/min) and k 2 (g/ (mg min)) are the adsorption capacity at equilibrium, the adsorption capacity at time t, PFO rate constant and PSO rate.
The results of the fitting are illustrated in Fig. 8b-c and Table 3. As shown in Fig. 8b-c, for Cu 2+ and Pb 2+ ions, the value of correlation coefficient (R 2 ) indicated a better fit to PSO model, and the Q e calc values obtained by pseudo-second order stand much closer with the experimental results Q e exp . These observations suggested that the overall rates of Cu 2+ and Pb 2+ ions adsorption were controlled by PSO kinetics, and chemisorption is the rate limiting step that controlled the adsorption process.

Table 2
Fitting parameters and errors of isothermal adsorption models (Langmuir, Freundlich and Temkin) for Cu2+ and Pb2+ ions onto 0.2APTES-SAAs

Thermodynamic parameters of adsorption process
In order to further understand the adsorption processes of Cu 2+ and Pb 2+ ions onto 0.2APTES-SAAs, thermodynamic parameters such as entropy change (ΔS°), enthalpy change(ΔH°) and Gibbs free energy change (ΔG°) in standard state were studied and calculated using the following Eqs. (9-12) (Oskui et al. 2019b;Gupta et al. 2019): where, R (J/ (mol K)), T (K) and K c are the gas constant, the absolute temperature and the distribution factor, respectively.
K c can be calculated by Eq. (12). ΔS° and ΔH° of adsorption can be calculated from the intercept and incline of the linear plot of lnK c against 1/T (Fig. 9), respectively. The determined thermodynamic adsorption parameters are summarized in Table 4. The negative value of ΔH°, ΔS° and ΔG° indicated that the adsorption was exothermic and spontaneous in nature with decreased randomness at the solid-liquid interface.

Adsorption mechanism
To further demonstrate the adsorption mechanism between AMSAAs and heavy metal ions, taking the adsorption of Cu 2+ ions onto 0.2APTES-SAAs as an example, XPS and SEM-EDS characterization were used to analyze and compare the microstructure changes of 0.2APTES-SAAs before and after Cu 2+ adsorption as demonstrated in Figs. 1 and 10. It can be seen from Fig. 10 that 0.2APTES-SAAs after Cu 2+ adsorption still remained a three-dimensional network structure composed of nanoparticles and Cu 2+ ions were uniformly distributed in the whole structure.  Moreover, as can be seen from Fig. 11a, the full XPS spectra of 0.2APTES-SAAs after Cu 2+ adsorption presented that Si, Al, O, C, N, Cu elements coexisted. The high-resolution spectrum of N 1 s could be split into three peaks (Fig. 11b): the peak at 399.6 eV corresponded to the C-N bond in the branched C-NH 2 structure of the modifier (Roy et al. 2020), the peak at 400.5 eV agreed well to the free state of the amino group and the peak at 402.2 eV corresponded to the amino group that was protonated or interacting with Cu 2+ (Shao et al. 2018). The peak of N element in the amino group shifted to the high binding energy due to that the N atom provided lone pair electrons to Cu 2+ , which resulted in the decreased electron cloud density of N, thus having a high binding energy (Zhang et al. 2016). In addition, the peak at 934.45 eV in Fig. 11c was ascribed to the characteristic peak of Cu 2p 3/2 , and the satellite peak at 942.99 eV further proved the existence of Cu 2+ (Gaudin et al. 2016) in the 0.2APTES-SAAs after Cu 2+ adsorption.
In summary, Cu 2+ was indeed adsorbed in 0.2APTES-SAAs structure and existed in the form of Cu 2+ . Since the N atom in 0.2APTES-SAAs structure had a lone pair of electrons and Cu 2+ had an empty orbital, a stable coordination bond could be formed between them (Xiao et al. 2019) as depicted in Fig. 12.

Recyclability and selectivity of 0.2APTES-SAAs
Four consecutive adsorption-desorption cycles of as-prepared 0.2APTES-SAAs were conducted, and the corresponding removal efficiencies for Cu 2+ ions were presented in Fig. 13. It can be seen that the removal efficiency of the absorbent decreased slightly from 98.8% (first cycle) to 87.5% (fourth cycle), indicative of excellent cycling stability of the AMSAAs. Moreover, in order to clarify the selective adsorption ability of AMSAAs for Cu 2+ and Pb 2+ ions, a ternary solution with Cu 2+ -Pb 2+ -K + ions was used as an objective. The concentrations of Cu 2+ , Pb 2+ and K + ions after adsorption given in Table S2 showed that Cu 2+ and Pb 2+ ion concentrations reduced significantly after adsorption, while that of K + Fig. 11 XPS spectra of 0.2APTES-SAAs after Cu 2+ adsorption (a) full spectrum, (b-c) high-resolution XPS spectra of N 1 s and Cu 2p regions  ions was basically unchanged, suggesting that the AMSAAs could selectively adsorb Cu 2+ and Pb 2+ ions in the solution. Table 5 summarized the comparison results of adsorption properties of Cu 2+ and Pb 2+ ions onto reported modified silica aerogels. The as-prepared AMSAAs from fly ash in this work for Cu 2+ and Pb 2+ ions possessed the best adsorption properties. This means that the simple preparation method proposed in this paper not only realized the use of fly ash with high added value but also had far-reaching significance to achieve the treatment of heavy metal ions in wastewater.

Conclusions
In this work, the amino-terminated SiO 2 -Al 2 O 3 composite aerogels used fly ash as raw material, APTES and AEAPMDS as modifiers were successfully synthesized by a sol-gel graftingatmospheric drying method and the APTES modified aerogel exhibited better adsorption properties for the shorter branched chain structure. Under the optimum conditions, 0.2APTES-SAAs achieved excellent adsorption properties for Cu 2+ and Pb 2+ ions as high as 195 mg/g and 500 mg/g within 10 min, respectively. Moreover, the adsorption processes of Cu 2+ and Pb 2+ ions by 0.2APTES-SAAs were more consistent with Freundlich model and also conformed to the pseudo-second-order kinetic model, which indicated that the adsorption processes of metal ions by AMSAAs were heterogeneous multilayer adsorption and controlled by the chemical reaction between the AMSAAs and heavy metal ions. The negative values of ΔS°, ΔH° and ΔG° confirmed the process of adsorption process was spontaneous and endothermic in nature with decreased randomness at solid-liquid interface. Moreover, the good cyclic stability of the adsorbents makes it possible to treat heavy metal ions in wastewater.
Author contribution MG and MZ contributed to the study conception and design. Material preparation, data collection and analysis were performed by YL, FY and HW. The first draft of the manuscript was written by FY, and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.
Funding This work was supported by the National Natural Science Foundation of China (U1810205) and the Shanxi Collaborative Innovation Center of High Value-added Utilization of Coal-related Wastes.

Data availability
The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Declarations
Ethics approval and consent to participate Not applicable.

Competing interests
The authors declare no competing interests.