1.   Introduction

Wastewater from chemical emissions, oil spillages, and biomedical pollutants poses a serious threat to humans and the environment. Hence, wastewater remediation is highly important. Toxic heavy metal ions (Cu²⁺, Pb²⁺, Cd²⁺, Zn²⁺, Hg²⁺, Ni²⁺, and ${\rm{Cr}}_{2}{\rm{O}}_{7}^{2-}$) are nondegradable and are carcinogenic to humans [1]. Therefore, heavy metal ions must be properly treated before being discharged to the environment. Conventional methods of wastewater treatment are filtration [2], precipitation [3], ion exchange [4], electrochemical treatments [5], and photocatalysis [6–7]. Despite their efficiency in the removal of heavy metal ions, these methods suffer from various limitations. For instance, the chemical precipitation routes show low removal efficiency and sometimes leave large amounts of sludge after use [3]. In addition, photocatalysis produces undesirable by-products, such as Cr(III) ions from Cr(VI) reduction, and thus needs a secondary treatment [6–7]. Meanwhile, ion exchange suffers from high pH dependencies [4]. Moreover, membrane filtration and electrochemical treatment are expensive to become commercially available [2,5]. With these considerations, absorption is still a cost-effective way to remove heavy metal ions (even at trace concentration). Other advantages of this method include easy implementation, simple design, none/less secondary products, effective cost, and being recyclable [8], enabling its wide applications.

To date, extensive absorbent materials, such as active carbon, zeolite [9], and organic resins [10], have been studied. However, active carbon, although cheap, is limited by its low absorption capacities and nonselectivity. The organic resin is almost nonrecyclable, nonregenerative, and thermally unstable [8,10]. Recent studies have reported that nanomaterials are promising in removing heavy metal ions [11–12] but are expensive and complex at processing [13]. Furthermore, they tend to agglomerate [12], which severely hinders their wide applications. Consequently, efficient absorbents with nanostructures and bulk structures are still highly in demand. Aerogels are excellent materials bridging nanostructures into intact bulk materials while preventing nanomaterial from aggregation [14]. Recently, Maleki has reviewed various aerogels [1], including inorganic [15–25], organic [26–27], carbon [28–31], semiconductor chalcogenide [32–33], and natural based aerogels [34–36], which have been used in environmental protection. Nanostructured Al₂O₃–SiO₂ shows high efficiencies in removing heavy metal ions because of its high surface area and rich surface functional groups containing oxygen [37]. However, retaining the intact bulk shape while maintaining the excellent dispersion of individual Al₂O₃/SiO₂ nanoparticles is challenging. On the one hand, the hydrolysis rate and pH value are challenging to adjust, especially when mixing two sols (e.g., SiO₂ and Al₂O₃ ones). This phenomenon could lead to severe phase separations and nonhomogeneity and thus poor absorption. A similar but simpler propylene oxide (PO)-assistant sol–gel method (without using various dry controlling agents, such as formamide [38–39]) could be employed to overcome the aforementioned problems. On the other hand, although supercritical [34] and freeze drying [35–36] have been explored for this purpose, the former is expensive and dangerous because of the high pressure involved [21], and the latter induces severe cracks. Consequently, a new drying method needs to be developed urgently. A recent study has shown that an organic solvent sublimation drying (OSSD) method [21] can keep the aerogel intact and retain the surface area by minimizing the capillary forces in the channels/pores of a wet gel [21].

In this work, we designed an aerogel structure to assemble the Al₂O₃/SiO₂ nanoparticles and to obtain high surface area by using an improved sol–gel and OSSD technique. The improvements are that almost no drying agents (as listed in reference [38–39]) are used, and the drying period can be shorten to 3 d. In this way, the dispersion (without agglomeration) of Al₂O₃–SiO₂ nanostructures is improved while the intact bulk aerogel structure is mostly maintained through a simplified PO-assistant sol–gel method and a facile OSSD drying technique. Furthermore, Cu²⁺, a common heavy metal ion in industrial wastewater and a major pollutant, is chosen as the model heavy metal ion to test the absorption properties of the aerogels in this work.

2.   Materials and methods

A clear Al₂O₃ sol was prepared through mixing 17 mL (0.3 mol) C₂H₅OH, 15.2 mL (0.86 mol) demineralized water, and 4.83 g (0.02 mol) AlCl₃·6H₂O at 50°C for 30 min under magnetic stirring. Similarly, a SiO₂ sol was prepared via reacting 0.56 mL (0.0025 mol) Tetraethyl orthosilicate (TEOS) with 0.18 mL (0.01 mol) demineralized water at 50°C for 90 min. The two sols were subsequently mixed uniformly, in which PO (0.16 mol) was added at 50°C for 10 min to allow the cross-linking reaction. The sol was changed to a wet gel, which was stored at room temperature for 2 h and aged in C₂H₅OH for 24 h. Subsequently, a gradual solvent exchange procedure was conducted in 3 d [21] (day 1: 100wt% C₂H₅OH, day 2: 50wt% C₂H₅OH and 50wt% acetonitrile, day 3: 100wt% acetonitrile), which is shorter than the usual 5-d scheme used in [21] and is an improved OSSD method. It was then dried in a vacuum chamber in accordance with the OSSD protocol at 80–100 Pa and 50°C under the same mechanism [21]. The as-dried gel was heated at 600–1200°C for 2 h in a muffle furnace before further characterizations.

The as-prepared sample was identified using X-ray diffraction (XRD) at 30 mA and 40 kV (XRD TTRIII, Rigaku, Japan) with Ni-filtered Cu K_α radiation. Its molecular structures and chemical composition were characterized using Fourier transform infrared (FTIR) spectrometry in transmission mode (SHIMAZDU spectrometer model IRAFFINITY-1, FTIR-8400S). The porosity, pore size distribution, specific surface areas, and absorption–desorption loops of the sample were measured using a QuantaChrome instrument (QuadraSorb SI). In addition, the microstructures of the sample were observed via a field-emission scanning electron microscopy (SEM, Zeiss SUPRA55, Germany) and a transmission electron microscope (TEM, JEM 2100, JOEL, Japan). Finally, an absorption experiment was performed where 100 mL of CuSO₄ solution containing (60 mg/L Cu²⁺) was used as the model solution, and its supernatant was taken out at various times after its contact with the as-prepared aerogel. The Cu²⁺ concentration of the supernatant was measured using an atomic absorption spectrometer (MKIIM6, Thermo, USA), and the Cu²⁺ concentration data were recorded and plotted with contact time.

3.   Results and discussion

3.1   Phase characterization

The XRD patterns (Fig. 1) revealed that the as-prepared samples contain alumina (JCPDS card No. [04-0875]) at 600 and 800°C and some minor amorphous or less crystallized phases appearing as a bump at the 2θ angles of 20°–30°. When the temperature was increased to 1000°C, the Al₂O₃ peaks became sharper and narrower, whereas the amorphous phase disappeared probably because it reacted with or doped into Al₂O₃ upon the heat treatment. When the temperature was further increased to 1200°C, mullite (JCPDS card No. [02-0431]) was detected. Therefore, the unidentified phase may be SiO₂ because a similar bump at the identical 2θ angles (20°–30°) in the corresponding XRD profiles was previously proven to be SiO₂ [22,38]. The existence of Si–O–Si or Si–OH in the network below 1000°C was then confirmed.

Figure  1.  XRD patterns of the as-prepared aerogels after firing at 600–1200°C.

下载: 全尺寸图片 幻灯片

3.2   FTIR observation

FTIR spectroscopy was performed on various samples, including the as-dried, 800°C-heated and 1000°C-heated samples, to verify the formation of Si–O–Si, Al–O–Si, or Si–OH bond in the as-prepared aerogels (Fig. 2). The as-dried one showed Si–OH stretching vibration at 956 cm⁻¹, confirming its existence in the gel network and verifying the reaction between TEOS and H₂O has taken place. When the temperature was increased to 800 and 1000°C, the Si–OH peaks (956 cm⁻¹) shifted towards 800 cm⁻¹, suggesting that the H atoms were gradually replaced by Si atoms in the network, forming the symmetric Si–O–Si bond [27]. This result proves that the Si–OH bond initially formed as early as in the drying stage and then transformed into Si–O–Si bond at later processing stages [38–39].

Figure  2.  FTIR of the as-prepared aerogels after drying and after firing at 800 and 1000°C (v is for stretching vibration).

下载: 全尺寸图片 幻灯片

On the other hand, Al–OH (474 cm⁻¹), Al–O (632 cm⁻¹), and Al−O−Al (1050 cm⁻¹, stretching vibration) were also detected, indicating the hydrolysis reaction between AlCl₃·6H₂O and PO has taken place [38–39]. Furthermore, some Al–O–Si (1069 cm⁻¹, stretching vibration [24]) bonds have also been detected when increasing temperature gradually. This result may be attributed to the condensation reactions between Al−OH and Si−OH/Si−OC₂H₅.

The presence of Al₂O₃ rather than SiO₂ in the XRD (Fig. 1) and FTIR (Fig. 2) results can be ascribed to the much faster gelation speed and larger composition quantity of the former than the latter [25]. In other words, as the Al−O−Al bond formed, the chance for Al–OH to react with Si–OH/Si–OC₂H₅ and form Al−O−Si bonds decreased. Hence, a mixture of mainly Al–O–Al and a possibly small amount of Si–O–Al was produced. Subsequently, Si–OH started to react with itself or Si−OC₂H₅, forming Si–O–Si. Consequently, SiO₂ and Si–O–Si were detected in the XRD and FTIR profiles (Figs. 1 and 2), respectively.

Other peaks, such as C–H (2990 cm⁻¹) and H–O–H (1610 cm⁻¹), were also observed in the as-dried and 800°C-heated samples, indicating that some organic solvents remained after the gelation/cross-linking process. When the temperature was increased to 1000°C, the –OH and C–H completely disappeared due possibly to the pyrolysis of these organics.

3.3   Microstructure

Fig. 3 shows that the as-prepared aerogels heated at 800°C are porous with pore diameters ranging from 10 to 20 nm and locating between the loosely packed nanoparticles (average diameters of 50–100 nm). The pores are quite homogeneous. Figs. 4(a) and 4(b) from two different locations both show that the as-prepared Al₂O₃–SiO₂ composite aerogels are needle-like and/or plate-like under a transmission electronic microscope, which is consistent with reference [40–41]. The Energy-dispersion X-ray spectroscopy (EDX) shows that the gels are composed of Al, Si, and O. This, together with the XRD and FTIR results, verifies the formation of Al₂O₃–SiO₂ composite aerogels. Element mapping (Figs. 3(d)–3(f)) shows that Al, Si, and O are presented homogeneously at identical positions in the aerogel. Overall, the PO-assisted sol–gel technique enabled the gelation better than traditional sol–gel techniques by avoiding the long pH adjustment procedures and controlling the hydrolysis rate [38–39].

Figure  3.  SEM (a, b) of the Al₂O₃–SiO₂ composite aerogel, EDX (c) of point 1 in (b), and element mapping (d–f) corresponding to (b).

下载: 全尺寸图片 幻灯片

Figure  4.  Transmission electron microstructure of the Al₂O₃–SiO₂ composite aerogel heat treated at 800°C at two different locations (a) and (b).

下载: 全尺寸图片 幻灯片

3.4   Specific surface area and porosity

As seen from the adsorption–desorption loops (Fig. 5(a)), some clues can be drawn about the pore shapes from its H2 typed loops. These pore shapes might be stacking between the uniform nanoparticles, which is consistent with the SEM observations (Fig. 3). Furthermore, narrow distributions of pore sizes (Fig. 5(b)) were observed with the meso-pore sizes. The average diameters were 9.7, 13.4, 14.8, and 15.8 nm for the as-dried, 600°C-heated, 800°C-heated, and 1000°C-heated samples, respectively (Table 1).

Table  1.  Average pore diameters, total pore volume, porosity, and specific surface areas of the as-dried, 600°C-heated, 800°C-heated, and 1000°C-heated samples

Sample	Average pore diameters / nm	Total pore volume / (mL·g−1)	Porosity	BJH specific surface area / (m2·g−1)	BET specific surface area / (m2·g−1)
As-dried	9.7	1.46	86.0%	827.544	602.462
600°C	13.4	1.28	83.6%	657.778	381.520
800°C	14.8	1.08	81.2%	455.839	292.720
1000°C	15.8	0.59	70.0%	225.984	148.990

下载: 导出CSV 

| 显示表格

Figure  5.  Adsorption–desorption loop (a) and pore size distribution (b) of the aerogel after drying and firing at 600, 800, and 1000°C. P and P₀ stand for pressure at measurement and at standard temperature, respectively; V stands for the volume of absorbed N₂; d stands for pore diameter.

下载: 全尺寸图片 幻灯片

Besides, Table 1 shows that the Barrett–Joyner–Halenda (BJH) specific surface area of the as-dried aerogel was 827.544 m²/g, which was higher than those at 657.778, 455.839, and 225.984 m²/g after heat treatment at 600, 800, and 1000°C, respectively. The Brunauer–Emmett–Teller (BET) specific surface area and the pore volume data are also shown in Table 1, suggesting that the aerogel sample has the highest total pore volume of ~1.46 mL/g after OSSD drying, dropping to 1.28, 1.08, and 0.59 mL/g after heat treatment at 600, 800, and 1000°C, respectively. The corresponding porosity was calculated as 86.0% for the as-dried sample, 83.6% for the 600°C-heated sample, 81.2% for the 800°C-heated sample, and 70.0% for the 1000°C-heated sample. These data indicate that the main component is Al₂O₃ (Fig. 1) and its density is ~3.9 g/cm³. The porosity of this aerogel is up to 86.0%.

Such high surface areas and porosities can be possibly attributed to the OSSD drying method [11]. OSSD reduces the surface tensions and capillary forces between the aerogel channels and walls better than other drying methods [40]. It can also help the aerogel achieve relatively higher specific surface area and pore volume compared with previously reported aerogels (Table 2).

Table  2.  Comparison between this work and references

Parameters	This work	References
Specific surface area	225.984 to 827.544 m2/g	233.12–481.79 m2/g [38] SiO2 457.82 m2/g and Al2O3 413.46 m2/g [21]
Pore volume	0.6–1.5 mL/g	0.3–1.89 mL/g [40]

下载: 导出CSV 

| 显示表格

3.5   Cu²⁺ uptake by the Al₂O₃−SiO₂ composite aerogel

Preliminary tests were conducted on the absorption effect of aerogels on Cu²⁺. In these spectra, the remaining concentration of Cu²⁺ dropped sharply within 40 min and reached ~0 afterward (Fig. 6(a)), and the removal efficiency can reach 99% in 40 min (Fig. 6(b)). In addition, the q_e (the absorption uptake at equilibrium) profile shows that the equilibrium Cu²⁺ uptake can reach ~69 mg/g (Fig. 6(c)). These results are better than those obtained in [11]. This high absorption efficiency can be possibly ascribed to the high surface area as shown in the BET and BJH specific surface area measurements (Fig. 5 and Table 1). Furthermore, it has easy accessible pores (Fig. 3) and high porosity (Table 1), which has greatly facilitated the diffusion of Cu²⁺ into bulk aerogel materials [11]. Finally, the multiple oxygen-containing functional groups (Al–OH, Si–OH, Al–O–Si, etc.) found in the composite aerogels (FTIR in Fig. 2) and the homogeneous distribution of Al, Si, and O (Fig. 3) might have afforded the aerogel excellent absorption properties [37]. The high surface area, easily accessible pores with high porosity, and oxygen-containing functional groups have synergistically promoted the absorption of Cu²⁺.

Figure  6.  Dependence of (a) Cu²⁺ concentration, (b) removal rate (%), and (c) uptake q_e on contact time.

下载: 全尺寸图片 幻灯片

Absorption kinetics was examined by performing the following calculations as previously described [11]. The absorption data of Cu²⁺ by the Al₂O₃−SiO₂ composite aerogels are consistent with the calculations from the pseudo second-order kinetics model (R² = 0.9944), which is much better than the pseudo first-order kinetics model (R² = 0.8829), as shown in Fig. 7. Therefore, the absorption process may follow the pseudo second-order kinetics and be controlled by the chemical absorption where valence forces and electron exchanges take charge [11].

Figure  7.  Fit between the experimental Cu²⁺ absorption data and (a) pseudo second-order kinetics or (b) pseudo first-order kinetics. t is for absorption time; q_t and q_e stand for the uptake amount of Cu²⁺ per gram of the absorbent at time t and at equilibrium, respectively.

下载: 全尺寸图片 幻灯片

The matches between the experimental and calculated data were evaluated using a normalized standard deviation ∆q (%) [42], to validate the pseudo second-order model. The ∆q (%) for the pseudo second-order model (82.55) fits more closely than the pseudo first-order one (148.83). Therefore, we speculated that the absorption of Cu²⁺ on the Al₂O₃–SiO₂ composite aerogel follows the pseudo second-order kinetics.

4.   Conclusion

Al₂O₃–SiO₂ composite aerogels were prepared using the PO-assisted sol−gel and OSSD methods. XRD and FTIR suggested that the aerogels are composed of mainly Al₂O₃ after firing below 1000°C. The aerogels have a high surface area of up to 827.544 m²/g and high porosity of 86.0%. The H2 absorption/desorption loop and the SEM observations suggested that the pores are homogeneously located between the Al₂O₃−SiO₂ composite nanoparticles. The aerogels can absorb 99% Cu²⁺ (60 mg/L) in ~40 min and reach the equilibrium uptake at ~69 mg/g. Calculations showed that the absorption of the aerogels follows pseudo second-order kinetics. Such high efficiency can be attributed to the high specific surface areas, easy accessible microstructure with high porosity, and rich oxygen-containing functional groups in the aerogel network.

Acknowledgements

This work was financially supported by the National Natural Science Foundation of China (No. 51602018), the Beijing Municipal Natural Science Foundation (No. 2154052), the China Postdoctoral Science Foundation (No. 2014M560044), and the Fundamental Research Funds for the Central Universities, China (No. FRF-GF-17-B7).