3.1 Characterizations
In Fig. 1, the patterns of X-ray diffraction (XRD) for SiO2, ZA, ZA/SiO2 were presented. The XRD results of all samples revealed no other phases, affirming the purity of all samples. The ZA and ZA/SiO2 samples display the same characteristic diffraction peaks as the Zn-Al layered double hydroxide [29, 30]; the diffraction peaks (003), (006), (009), (012), (015), (110) and (113) at 10.04o, 20.02o, 30.38o, 33.12o, 33.48o 60.44o and 61.55o were shown, indicated ZA and ZA/SiO2 were prepared successfully via the hydrothermal treatment. While SiO2 only has a wide peak near 23o, demonstrating its amorphous morphology [27, 28]. The addition of SiO2 did not change the layered structure of Zn-Al layered double hydroxide, but only acts as a supporting template. At the same time, it increased its specific surface area, which would be conducive to the improvement of adsorption performance (see the section of 3.2).
Field emission scanning electron microscope (SEM) was utilized to detect the microscopic morphology of samples, and the result was showed in Fig. 2. The Fig. 2a illustrate that the silica spheres, measuring approximately 300 nm in size, exhibit a smooth surface. In the absence of SiO2 templates, the synthesized ZA sample consists of self-assembled nanosheets that form an aggregated spherical structure (Fig. 2d). It can be seen from the TEM (Fig. 2b) and SEM (Fig. 2c) images of ZA/SiO2 that with the addition of SiO2, a part of the silica pellets was wrapped in thin sheets, forming a core-shell structure. The SiO2 pellets, which were not coated by sheets, supported the smaller pores in the ZA in the outer layer, making the structure bulkier, which was consistent with BET analysis. Simultaneously, this phenomenon contributed to the augmentation of the specific surface area of ZA/SiO2 composites. This indicates that the spatial structure of ZA/SiO2, which were acquired through SiO2 loading, has been effectively optimized.
Fourier transform infrared spectroscopy (FT-IR) could be used to detect the surface functional groups of samples and to analyze the interactions between adsorbent and adsorbate molecules. Figure 3 shows the FT-IR spectra of SiO2, ZA, ZA/SiO2. In the spectra of each sample, the wide absorption peaks of 3200 ~ 3400 cm− 1 were the stretching vibration peak of the -OH on the laminate and the water molecule -OH between the laminate; the absorption peaks around 1400 cm− 1 were the vibration absorption of interlayer CO32− [28] The low frequency absorption peaks between 400 and 900 cm− 1 were the stretching and bending vibration peak of M-O and M-OH.
Table 1
The specific surface area (SBET), particle size distribution and average particle size values of SiO2, ZA and ZA/SiO2 samples.
Sample
|
SBET
(m2·g− 1)
|
Dx(10)
(µm)
|
Dx(50)
(µm)
|
Dx(90)
(µm)
|
D [3,2]
(µm)
|
D [4, 3]
(µm)
|
SiO2
|
13.3
|
2.87
|
16.3
|
45.8
|
7.54
|
23.4
|
ZA
|
62.4
|
2.80
|
16.5
|
56.4
|
7.67
|
29.1
|
ZA/SiO2
|
62.5
|
2.25
|
8.94
|
76.5
|
5.11
|
28.1
|
Figure 4 shows the N2 adsorption-desorption curves and pore size distribution curves of SiO2, ZA and ZA/SiO2 samples, respectively. It could be seen that the adsorption curves and desorption curves of all samples both form a closed curve, which was called the H3 type adsorption hysteresis ring, and its appearance further proves that SiO2, ZA and ZA/SiO2 samples are porous structures formed by the accumulation of flaky particles [31]. Table 1 shown the characterization results of the specific surface area (SBET), particle size distribution and average particle size values of SiO2, ZA and ZA/SiO2 samples. It was worth noting that the SBET values of ZA and ZA/SiO2 were very similar. As can be seen from Fig. 4b, compared with ZA and SiO2, the particle size distribution of ZA/SiO2 is relatively concentrated, indicating that the particle size distribution is uniform. The appearance of SiO2 provides a landing point for the lamellar structure of ZA, making it grow evenly. Generally, the smaller the particle size of the adsorbent, the larger the adsorption specific surface energy. And the particle size distribution is concentrated, indicating that the prepared adsorbent is more stable [31]. DX (10), DX (50) and DX (90) can reflect the uniformity of the powder sample, which is called the cumulative distribution or frequency distribution of the powder. Where DX (50) is called the median diameter or median particle size, it is an important parameter to measure the particle size distribution, representing the corresponding particle size when the cumulative particle size distribution percentage reaches 50%. A smaller DX (50) value means that the particle size is more evenly distributed and the particle size is smaller. From Table 1, it shown that the particle size of ZA/SiO2 (8.94 µm) is the most uniform, and the particle size of ZA (16.5 µm) and SiO2 (16.3 µm) is similar. Both D [3, 2] and D [4, 3] represent the average particle size on the basis of volume. Where D [3, 2] is called "volume area mean diameter", referred to as the area mean diameter, which is the ratio of the total volume of particles to the total area. D [4, 3] is called "mass moment volume mean diameter" and referred to as volume mean diameter. Both of them are the average diameters calculated from particle size distributions on different benchmarks. D [3, 2] is calculated by area distribution and D [4, 3] is calculated by volume distribution. It can be seen from Table 1 that ZA/SiO2 has the smallest average particle size by volume area, but its average particle size is similar to ZA.
3.2. Adsorption performance of samples
The kinetic study is one of the main reasons to reveal the adsorption reaction rate. Figure 5 respectively shown the adsorption kinetics curves of Cu(II) and Cr(Ⅵ) on ZA/SiO2 samples with different solid content. The adsorption rates of ZA/SiO2 to Cu(II) and Cr(Ⅵ) were fast in the first 30 min, and the adsorption equilibrium states were reached in about 180 min. This was because with the progress of the reaction, the adsorption site on the adsorbent was continuously occupied and gradually reaches a near saturation state. What’s more, it can be seen that under each solid content, the adsorption kinetics curves of ZA/SiO2 samples were L-shaped, and the adsorption capacity decreased with the increase of solid content, which indicates that there is an obvious effect of adsorbent concentration in the adsorption system, that is, the adsorption capacity decreases with the increase of adsorbent concentration [32, 33]. The sorption kinetics of Cu(II) and Cr(Ⅵ) on ZA/SiO2 samples were fitted with the pseudo-first order model and pseudo-second order model (Fig. 5a, b, a′ and b′), and the fitting parameters are listed in Table 2. At different solids content, adsorption kinetics data of ZA/SiO2 samples were well fitted by a pseudo-second-order kinetic model, indicating that chemical adsorption played a leading role in the adsorption process of Cu(II) and Cr(Ⅵ) by ZA/SiO2 [34].
Table 3
Nonlinear-fit data of model parameters for Cu(II) and Cr(VI) sorption on SiO2, ZA, ZA/SiO2.
Adsorption object
|
Adsorbents
|
Langmuir isotherm:
qe = KLqmCe/(1 + qmKL)
|
|
Freundlich isotherm:
qe = KFCenF
|
qm
(mg·g− 1)
|
KL (L·mg− 1)
|
R2
|
nF
|
KF (mg1 − nFLnF·g− 1)
|
R2
|
Cu(II)
|
SiO2
|
8.82
|
0.0185
|
0.977
|
|
0.339
|
1.09
|
0.926
|
ZA
|
43.6
|
0.0432
|
0.923
|
|
0.223
|
11.0
|
0.998
|
ZA/SiO2
|
158
|
0.0137
|
0.959
|
|
0.592
|
5.35
|
0.995
|
Cr(VI)
|
SiO2
|
16.8
|
0.00432
|
0.892
|
|
0.701
|
0.176
|
0.831
|
ZA
|
98.6
|
0.00366
|
0.980
|
|
0.729
|
0.835
|
0.978
|
ZA/SiO2
|
176
|
0.00644
|
0.989
|
|
0.551
|
4.96
|
0.966
|
The adsorption isotherm would describe the relationship between the adsorption capacity and the equilibrium concentration of the adsorbent solution at 25°C and pH = 5.5. Figure 6 respectively shown the adsorption isotherms of Cu(II) and Cr(Ⅵ) on SiO2, ZA, ZA/SiO2 with 1 g·L− 1 solid content. It can be seen that the adsorption amount of ZA/SiO2 was the largest, followed by ZA, and the adsorption capacity of SiO2 was the smallest. Both Langmuir model and Freundlich model could fit the adsorption data well. The nonlinear-fit data of model parameters for Cu(II) and Cr(VI) sorption on SiO2, ZA, ZA/SiO2 was shown in Table 3 and the linear-fit data was shown in Table 4. As can been seen, the equilibrium adsorption capacity of ZA/SiO2 were 158 mg·g− 1 for Cu(II) and 176 mg·g− 1 for Cr(VI) from nonlinear-fit, which were 3.6 and 1.8 times of ZA, respectively. However, the equilibrium adsorption capacity of ZA/SiO2 were 208 mg·g− 1 for Cu(II) and 233 mg·g− 1 for Cr(VI) from linear-fit. Under both fitting conditions, the adsorption effect of ZA/SiO2 on Cr(VI) is better than that of Cu(II).
Table 4
Linear-fit data of model parameters for Cu(II) and Cr(VI) sorption on SiO2, ZA, ZA/SiO2.
Adsorption object
|
Adsorbents
|
Linear form of Langmuir isotherm:
\(\:\frac{{C}_{e}}{{q}_{e}}=\frac{{C}_{e}}{{q}_{m}}+\frac{1}{{K}_{L}{q}_{m}}\)
|
|
Linear form of Freundlich isotherm:
lg qe = lg KF + nF lg Ce
|
qm
(mg·g− 1)
|
KL
(L·mg− 1)
|
R2
|
nF
|
KF
(mg1 − nFLnF·g− 1)
|
R2
|
Cu(II)
|
SiO2
|
9.04
|
0.017
|
0.986
|
|
0.423
|
1.40
|
0.858
|
ZA
|
43.4
|
0.037
|
0.898
|
|
0.218
|
11.3
|
0.996
|
ZA/SiO2
|
208
|
0.008
|
0.964
|
|
0.640
|
4.27
|
0.990
|
Cr(VI)
|
SiO2
|
19.1
|
0.003
|
0.902
|
|
1.18
|
0.014
|
0.866
|
ZA
|
120
|
0.003
|
0.576
|
|
0.720
|
0.855
|
0.932
|
ZA/SiO2
|
233
|
0.004
|
0.944
|
|
0.739
|
2.04
|
0.977
|
3.3 DFT analysis of Cu(Ⅱ) and Cr(Ⅵ) removal mechanism by ZA/SiO2
The distinct ionic radii of Cu (II) and Cr (VI) give rise to inherent variations in the adsorption process of ZA/SiO2 for these two ions. Therefore, the adsorption mechanism at electronic scale was explored in-depth by density functional theory.
Figure 7a, b, and c depict that when the Zn atom was replaced with Cr(VI), the electrons lost by Cr(VI) were transferred to the neighboring oxygen atom. In comparison to the pre-replacement model, a greater proportion of the chemical bonds formed by Cr(VI) with the surrounding O atoms assumed the form of ionic bonds. This observation suggests that Cr(VI) has a propensity to preserve its positive charge by relinquishing its own electrons. Consequently, Cr(VI) was drawn towards the surrounding electronegative O atoms, where it existed in a stable state. Simultaneously, Cr(III) formed chemical bonds with adjacent O and H atoms. Moreover, taking into account the extent of the electron cloud, it can be deduced that electrons from Cr(III) were primarily moved to the hydroxyl atoms within the interlayer, leading to the formation of a stable Cr(OH)3 precipitate. Additionally, the elimination of Cr(VI) from water might also be linked to the incorporation of Cr2O72−.
Figure 7c and d depict the substitution of the Zn atom by Cu(II), where the transfer of electrons occurred towards the neighboring O atoms. However, in the case of Cu(II), the quantity of transferred electrons was smaller compared to what was observed for Cr(VI), and the ratio of ionic bonds following Cu(II) substitution was A bit greater than that in the case of Cr(VI). Despite the presence of an electron cloud surrounding Cu(II), the most extensive area was observed near the oxygen (O) atoms adjacent to Zn(II) (as shown in Fig. 7c), indicating a predominant transfer of electrons to the O atoms of Zn-O. This presents a contrasting electron transfer pathway compared to the ion exchange process observed in Cr(VI) removal. Additionally, the energy needed for Cu(II) removal via ion exchange was less than that required for Cr(VI) removal. This distinction can be attributed to the notable difference in the ionic radius of Cu(II), which was considerably smaller compared to that of Cr(VI) and was closer in size to that of Al(III). Therefore, not only did the Cu(II) ion exchange process have a lower energy requirement, but it also displayed superior structural stability in comparison to Cr(VI).
The complexation model for Cu(Ⅱ) and Cr(VI) adsorption by ZA/SiO2 was optimized. Figure 7e illustrates the adsorption energy calculated post-structural optimization, where Eads = − 2.12 eV and − 3.27 eV for ZA/SiO2–Cu (Ⅱ) and Cr(VI) complex form and Eads <0. The results show that under mild condition, ZA/SiO–Cu(Ⅱ) and Cr(VI) complexation cannot occur spontaneously. The deduction can be made that the -OH groups within ZA/SiO- were involved in the complexation of Cu(Ⅱ) and the removal of Cr(VI). To acquire a more comprehensive understanding of the interfacial interactions, the d-band centers (εd) and projected d-band densities of Cu(II) and Cr(VI) atoms at their stable adsorption locations in pristine ZA/SiO2 with hydroxide interfacial layer slabs were calculated. Figure 7f illustrates that the (εd) of Cu(II) in pure ZA/SiO2 was lower than the (εd) of Cr(VI) in ZA/SiO2 with hydroxide interfacial layer slabs. Consequently, the adsorption affinity of Cr(VI) on the surface of ZA/SiO2 with a hydroxide interface layer was stronger than that of pure ZA/SiO2. This was primarily due to the interface interaction, which mainly involves the modification of electron transfer and the shift in the D-band center.
Upon conducting comprehensive characterization analysis and batch adsorption experiments, it becomes evident that the adsorption of Cu(II) and Cr(VI) by ZA/SiO2 entails a multifaceted process marked by diverse mechanisms.
ZA/SiO2 possess an exceptional porous structure and a significantly expanded specific surface area, which provides numerous effective adsorption sites within the pore channels for Cu(II) and Cr(VI) uptake. The binding of Cu(II) and Cr(VI) to ZA/SiO2 may result from multilayer physical adsorption. Regarding the saturated adsorption capacity (qm) of ZA-LDHs for pollutants, the respective adsorption capacities for each contaminant can be ranked as follows: Cr(VI) > Cu(II). In summary, the primary mechanism for removing Cr(VI) from water involves isomorphic substitution of Cr(VI) and Zn(II), as well as the intercalation of Cr2O72−, with the exception of Cr(OH)3 precipitation.
The preceding phenomenon can be explained by the smaller size of Cr(VI) (0.044 nm) compared to Zn(II) (0.074 nm), whereas the latter can be understood as a result of the successful incorporation of SiO2 (0.86 nm × 0.59 nm × 0.18 nm) at the initial stages, thus creating a pathway for the entry of Cr2O72− (0.55 nm × 0.25 nm × 0.21 nm). Regarding the primary mechanism for eliminating Cu(II) from water, it involves isomorphic substitution as the predominant process, except for the formation of Cu(OH)2 precipitates. This phenomenon can mainly be attributed to the very similar ionic radii of Cu(II) (0.073 nm) and Zn (II) (0.074 nm).