Abstract

Geopolymerization is a developing reaction process for the utilization of solid wastes. In the present study, fly ash-based geopolymer and its derivative (Fe(II)-modified geopolymer) were synthesized and characterized using XRD, SEM, FTIR, BET, UV-Vis DRS as well as TG-DTA, and adopted as adsorbents for removal of Cs+ and Sr2+, and from solutions. Each sorption kinetic was well fitted to the pseudo-second-order model. The sorption of Cs+ and Sr2+ onto original geopolymer were better fitted to the Langmuir model. However, the Freundlich model is more befitting for sorption of onto Fe(II)-modified geopolymer. The free energies calculated from the D-R isotherm indicated that the sorption for Cs+ and Sr2+ were dominantly ion exchanges. Ring size plays a decisive role in ion exchanges for both Cs+ and Sr2+. Furthermore, the arrangement of SiO4 and AlO4 tetrahedrons has significant impacts on the ion exchange of Sr2+. XPS results indicated that a part of Fe2+ in Fe (II)-modified geopolymer had been oxidized to Fe3+ after sorption. Precipitation of FeAsO4 could partially contribute to the arsenate removal from solution. sorption has also occurred through the formation of inner-sphere complexes via ion exchange reaction, which could be predominantly attached by bidentate linkages.

INTRODUCTION

Geopolymers are a class of amorphous aluminosilicates with a three-dimensional structure consisting of SiO4 and AlO4 tetrahedrons, usually synthesized by low-temperature polycondensation of different materials such as metakaolin, and coal fly ash (Rożek et al. 2018). They always exhibit high compressive strength, high-temperature resistance, and permeability resistance, which provides tremendous potential to be a desirable alternative to ordinary Portland cement (Cho et al. 2017). Generally, the cations, such as Na+ and K+, would be located in the cavities of the framework in order to balance the negative charge produced by the AlO4 tetrahedron. These charge-balancing cations can be exchanged with other cations, suggesting that geopolymer can be used as an adsorbent like zeolite in the removal of heavy metals from aqueous solutions. Water pollution by heavy metals is always a serious problem, which has been in the limelight for so many years (Uddin 2017). Currently, there are many approaches referring to heavy metal wastewater treatment, including adsorption, chemical precipitation, membrane filtration, ion exchange, and electrochemical treatment. Among them, adsorption, presenting the advantages of high efficiency, low cost, and a simple operation process, has been considered to be one of the most attractive options. The removal of Cu2+, Pb2+, Zn2+, Mn2+, Co2+, Ni2+, Cs+, etc. by adsorption onto geopolymer has been conducted, resulting in good performances (El-Eswed 2018). However, geopolymer has a low affinity with anions, such as , (Al-Mashqbeh et al. 2018), because of the permanent negative charge on the AlO4 tetrahedron. A chemical modification would be needed to apply geopolymer for anion removal. It has been shown that can be effectively removed by barium-modified blast-furnace-slag geopolymer (Runtti et al. 2016).

Coal fly ash is an industrial waste composed of the fine particles that are captured from flue gases by electrostatic precipitators or other particle filtration equipment. Every year, large amounts of fly ash are generated over the world, which has been a serious environmental problem (Tian et al. 2019). Hence, synthesis of geopolymer could be a good way to consume coal fly ash and, meanwhile, the geopolymer could be used for other applications, which has great significance for the environment and economy. In order to evaluate the ability of fly ash-based geopolymer in adsorption of heavy metals, Cs+, Sr2+ and were selected in this study. Cs+ and Sr2+ generally appear as radionuclides in nuclear waste effluents, so stabilization of them is required. Since they possess positive charges, it is appropriate to utilize geopolymer as an adsorbent directly. However, as discussed above, there is a low affinity between geopolymer and anion. Therefore, Fe(II)-modified geopolymer was developed for the removal of in this research. The detailed reaction mechanisms for these sorption processes are discussed and proposed.

EXPERIMENTAL

Materials

The coal fly ash used in this study was collected from a thermal power plant in China. The X-ray diffraction (XRD) pattern (Figure 1) shows that mullite and quartz are the main crystal minerals in the sample. The chemical composition of fly ash was determined using X-ray fluorescence spectroscopy (XRF, Rigaku ZSX Primus II). The result showed that the SiO2 and Al2O3 contents in fly ash were 44.60 and 46.10%, respectively. The synthesis procedure of the geopolymer was as follows: the module of sodium silicate (SiO2/Na2O) was adjusted to 1.2 by adding sodium hydroxide. Then the activated solution was mixed with fly ash at a mass ratio of 0.67. After stirring for 10 min, the slurry was put into a mold with a diameter of 30 mm and thereafter treated by ultrasonication for 5 min in order to remove bubbles. The specimen was sealed with a plastic bag and cured at 80 °C for 12 h. Then it was cured at room temperature for another 18 h. The paste was then crushed and ground, then sieved to collect under 150 mesh sieve. Washing was conducted to remove the extra alkali. Lastly, the powder was dried at 105 °C overnight.

Figure 1

(1) XRD patterns of raw fly ash, geopolymer and Fe(II)-modified geopolymer. Symbols: ◇, mullite (3Al2O3·2SiO2); ○, quartz (SiO2); (2) SEM images of (a) coal fly ash, (b) geopolymer and (c) Fe(II)-modified geopolymer.

Figure 1

(1) XRD patterns of raw fly ash, geopolymer and Fe(II)-modified geopolymer. Symbols: ◇, mullite (3Al2O3·2SiO2); ○, quartz (SiO2); (2) SEM images of (a) coal fly ash, (b) geopolymer and (c) Fe(II)-modified geopolymer.

In addition, Fe(II)-modified geopolymer was prepared as followed: 5 g of geopolymer was put in 1 L of 0.1 mol·L−1 FeSO4 with stirring under N2 protection. After 20 h, the obtained solid was washed several times with deionized water that had dissolved O2 removed through N2 bubbling. Then it was dried on a freeze dryer for 24 h and stored under N2 protection. In addition, simulated pollutants Cs+, Sr2+ and were prepared by dissolving proper amounts of CsCl (special grade, Wako), SrCl2·6H2O (special grade, Wako) and KH2AsO4 (90%, Wako) in deionized water, respectively.

Sorption experiments

50 mg of geopolymer or Fe(II)-modified geopolymer was taken and put in the 50 mL metal ion solution in variable concentrations at room temperature. Effects of solution pH and the concentration of adsorbent on the removal efficiencies were explored and the results are shown in Figures S1 and S2 (available with the online version of this paper). Thus, based on these results, experimental conditions were definite for kinetic and isotherm studies. The initial pH was adjusted to 7 for adsorbing Cs+ and Sr2+, and 5 for sorbing , respectively, by adding 0.1 mol·L−1 HCl or NaOH. Then the solutions were put on a magnetic stirrer for 24 h in order to reach equilibrium. Lastly, the 0.2 μm filter was used to remove fine particles from the solution for subsequent concentration determination. The amount of sorbed ions (qe) was calculated by the equation below. 
formula
(1)
where Ce and C0 are the equilibrium and initial concentrations of metal ions (mmol·L−1), V is the volume (L) and m is the weight of the adsorbent (g).

Kinetic studies were performed using the initial concentration of 100 mg/L in the case of Cs+ and Sr2+ sorption onto the original geopolymer. For the , the initial concentration was 50 mg/L. For these investigations, 100 mg of adsorbent was added to beakers containing 100 mL solution with target pollutants. Then, the mixture was stirred using a magnetic stirrer at room temperature with an initial pH of 7 for Cs+ and Sr2+, and 5 for . Set volumes (1 mL) of aliquot were collected as a function of the contact time, while the solution was continuously stirred. The adsorbent in the collected solution was removed using a 0.2 μm filter.

Characterizations

The concentrations of Sr and As were determined using inductively coupled plasma optical emission spectrometry (ICP-OES, Perkin Elmer, Optima 8300, USA) and Cs was measured using inductively coupled plasma mass spectrometry (ICP-MS, Agilent 7500ce, USA). The XRD patterns were obtained at room temperature on a Rigaku Ultima IV XRD (Akishima, Japan): Cu Kα (40 kV, 40 mA) with a Ni filter at a scanning speed of 2° min−1 and scanning step of 0.02°. The Fourier transform infrared (FTIR) spectra (400–4,000 cm−1) were recorded using an FTIR spectrometer (JASCO 670 Plus, Japan) with a resolution of 4 cm−1 employing powder samples diluted using KBr. The specific surface areas of raw fly ash and adsorbents were determined by a high-precision surface area and pore size distribution analyzer (BEL-Max, BEL, Japan) at −196 °C. Pre-treatment under vacuum at 100 °C for 15 h was conducted in order to remove adsorbed gas and water. In addition, thermogravimetric-differential thermal analysis (TG-DTA, 2000 SA thermal balance, Bruker, Germany) was used to determine the thermal properties of the adsorbent using α-Al2O3 as the reference material. The heating rate and nitrogen flow were 10 °C/min and 120 mL/min, respectively. Diffuse reflectance spectroscopy (DRS) was carried out in the range of 200–800 nm on a UV-2450 spectrophotometer equipped with a diffuse-reflectance attachment, using BaSO4 as a reference. The dehydrated sample was prepared by calcining the original sample at 300 °C for 0.5 h with a heating rate of 10 °C/min and nitrogen flow of 120 mL/min. Scanning electron microscopy-energy dispersive X-ray spectroscopy (SEM-EDX) results were obtained from a VE-9800 SEM (Keyence, Osaka, Japan) with 20 kV acceleration voltages. X-ray photoelectron spectroscopy (XPS) was used to determine the chemical state of Fe in Fe(II)-modified geopolymer before and after adsorption of , performed on an ESCA 5800 (ULVAC-PHI, Inc., Kanagawa, Japan) using a monochromated Al Kα X-ray source. Here, the binding energy of C1s = 284.6 eV for adventitious carbon was adopted to calibrate the binding energies.

RESULTS AND DISCUSSION

Characterization of adsorbents

XRD patterns and SEM images of coal fly ash, original geopolymer and Fe(II)-modified geopolymer are shown in Figure 1. The main difference between them was the shift of the peak top of broad signals derived from the amorphous phase from approximately 23° for coal fly ash to almost 27° for geopolymer, indicating that the aluminosilicate glass phase in geopolymer was highly disordered. On the other hand, the XRD pattern of Fe(II)-modified geopolymer had no change compared to the original geopolymer. SEM images also indicated that the geopolymerization product was formed with the comparison of the smooth spheres of raw fly ash. Even though a relative amount of fly ash still did not completely dissolve, the geopolymer had been formed and overlapped on the surface of fly ash particles. This is the pivotal point to use geopolymer and its derivative as adsorbents.

The infrared spectroscopic results for raw fly ash, geopolymer and Fe(II)-modified geopolymer are presented in Figure 2(a). There was a marked difference between the spectra of fly ash and geopolymer at around 3,463 and 1,657 cm−1 which were assigned to stretching and bending modes of the water molecule, respectively. In the spectrum of fly ash, there is no corresponding peak because fly ash does is not able to ion exchange, so the water molecule cannot be bound with metal ions as in geopolymer or its derivative. The intense peak at 1,114.2 cm−1 was attributed to the asymmetric stretching vibration mode of T–O–Si (T = Si or Al). For geopolymer, the band of T–O–Si were shifted to 1,058.2 cm−1, which meant the rearrangement of SiO4 and AlO4 tetrahedrons. The peak at 580.6 cm−1 corresponded to the asymmetric stretching vibration of Si–O–Al and the band at 486.8 cm−1 was connected with the bending vibration O–Si–O in silicate tetrahedron. Blue shift and red shift occurred to these two bands, respectively. In addition, the stretching and bending vibrations of water molecules at 3,463.4 and 1,664.5 cm−1 in the spectrum of original geopolymer were shifted to 3,438.8 and 1,654.1 cm−1 for Fe(II)-geopolymer. This should be because Fe2+ has the larger polarizing power (charge/radius) and then stronger hydrogen bonds between the water molecules near Fe2+ would be formed (Madejová et al. 2002). Therefore, it can result in the bond length of O-H becoming longer, thereby leading to the wavenumber shifts of stretching vibration of them. In terms of other bonds including Si–O–Si, Si–O–Al, and O–Si–O, there were also some shifts that occurred to them. This was mainly because Fe2+ can be exchanged with Na+ in geopolymer, and this affected the local coordination chemistry.

Figure 2

(a) FTIR spectra of coal fly ash, geopolymer and Fe(II)-modified geopolymer; (b) UV-vis DRS spectra of dehydrated original geopolymer and Fe(II)-modified geopolymer; (c) TG-DTA curves of geopolymer and modified geopolymer.

Figure 2

(a) FTIR spectra of coal fly ash, geopolymer and Fe(II)-modified geopolymer; (b) UV-vis DRS spectra of dehydrated original geopolymer and Fe(II)-modified geopolymer; (c) TG-DTA curves of geopolymer and modified geopolymer.

Addition of the affinity with the anion to the geopolymer for anion can be an important modification in the stabilization/solidification of hazardous wastes. In the current study, the modification of geopolymer by Fe2+ was conducted and the obtained product presented a pale green color. The UV-Vis DRS spectra of original geopolymer and Fe(II)-geopolymer are shown in Figure 2(b). Compared to the original geopolymer, Fe(II)-geopolymer has a greater increase in adsorption in both the range of UV and visible light. Those bands at approximately 200–300 nm were assigned to the Fe-O charge-transfer transitions of Fe2+ in a tetrahedral and/or octahedral environment of oxygen atoms, which indicated that single Fe ions were present in cationic sites in the geopolymer.

TG curves of the original geopolymer and Fe(II)-modified geopolymer are presented in Figure 2(c). The weight loss at a temperature lower than 100 °C could be attributed to the loss of external weakly bound water molecules. The weight loss in the range of 160–330 °C is due to water molecules being more strongly bound to the mobile non-framework cations that are located in geopolymer cavities (Benaliouche et al. 2015). Greater weight loss happened in the case of Fe(II)-geopolymer (2.41%) than that of the original geopolymer (1.62%) at a temperature of 160–330 °C. It is obvious that cations with higher charge will interact more strongly with the unshared electron pair of the oxygen atom in the H2O molecule. Hence, the formation of the hydrogen bond between one of the hydrogen atoms of the water molecule and the nearest framework oxygen will be favored when the cation with higher charge density acts as a charge balancing cation. On the other hand, Fe2+ has a smaller radius (0.076 nm) than that of Na+ (0.095 nm), which indicates that there is more space for bound water in the cavities of geopolymer.

Nitrogen adsorption/desorption isotherms of raw fly ash, original geopolymer and Fe(II)-modified geopolymer are illustrated in Figure S4 (available with the online version of this paper). The specific surface area of raw fly ash was only 1.1 m2/g according to the Brunauer–Emmett–Teller (BET) calculation of isotherm. However, the specific surface area of original geopolymer increased up to 131.4 m2/g. As can be seen, a steep rise at low relative pressure (P/P0 < 0.01) occurred in the adsorption curves of the original geopolymer, implying that micropore was produced after geopolymerization of coal fly ash. As for Fe(II)-geopolymer, the steep rise at low relative pressure (P/P0 < 0.01) became lower compared to the original geopolymer, suggesting that the number of micropores decreased, thereby leading to the decrease of the specific surface area to 107.9 m2/g. the influence on the surface area by multivalent cations is relatively complex. This behavior may be explained by the fact that the pores in geopolymer can be partially blocked by the cluster when Fe2+ is in large numbers at sites inside the cages (Lopes et al. 2015).

Sorption kinetics and isotherms

To describe the changes in the sorption of metal ions with time, the linearized forms of pseudo-first-order and pseudo-second-order kinetic equations are given as following Equations (2) and (3) (Lin & Wang 2009). 
formula
(2)
 
formula
(3)
where qe and qt (mmol/g) are the amounts of metal ions sorbed onto geopolymer or Fe(II)-geopolymer at equilibrium and at time t, respectively, and k1 and k2 are the rate constants (min−1) of the pseudo-first-order and pseudo-second-order, respectively.

The fitting plots are shown in Figure S5 (available online), and the obtained parameters of each plot are summarized in Table 1. In case of the pseudo-second-order model, the correlation coefficient values for sorption of Cs+ and Sr2+ onto the original geopolymer and the sorption of onto the Fe(II)-geopolymer were higher than the values for the pseudo-first-order. In addition, the calculated values of qe for Cs+ (0.346 mmol/g), Sr2+ (0.242 mmol/g) and (0.129 mmol/g) based on the pseudo-second-order model had good agreements with the experimental values of qe for Cs+ (0.351 mmol/g), Sr2+ (0.253 mmol/g) and (0.110 mmol/g). Therefore, it can be reasonably concluded that the adsorption kinetics of the systems including sorption of Cs+ and Sr2+ onto original geopolymer and the sorption of onto Fe(II)-geopolymer fitted better with the pseudo-second-order model. This indicates that the overall rate constant of each sorption process appeared to be controlled by the chemical sorption.

Table 1

Calculated parameters of the pseudo-first-order and pseudo-second-order kinetic models for the sorption of Cs+ and Sr2+ onto original geopolymer, onto Fe(II)-geopolymer

Kinetic modelParameterCs+Sr2+
Pseudo-first-order qe, exp (mmol/g) 0.351 0.253 0.110 
qe, cal (mmol/g) 0.338 0.234 0.108 
k1 (min−13.306 1.461 0.030 
R2 0.985 0.954 0.967 
Pseudo-second-order qe, cal (mmol/g) 0.346 0.242 0.129 
k2 (min−123.173 12.509 0.266 
h (mmol g−1 min−12.774 0.733 0.004 
R2 0.998 0.977 0.983 
Kinetic modelParameterCs+Sr2+
Pseudo-first-order qe, exp (mmol/g) 0.351 0.253 0.110 
qe, cal (mmol/g) 0.338 0.234 0.108 
k1 (min−13.306 1.461 0.030 
R2 0.985 0.954 0.967 
Pseudo-second-order qe, cal (mmol/g) 0.346 0.242 0.129 
k2 (min−123.173 12.509 0.266 
h (mmol g−1 min−12.774 0.733 0.004 
R2 0.998 0.977 0.983 
In this study, isotherm models including Langmuir, Freundlich, and Dubinin-Radushkevich (D-R) were used to fit the equilibrium data for the sorption of Cs+ and Sr2+ onto geopolymer and the sorption of onto Fe(II)-geopolymer. The linearized forms of Langmuir and Freundlich are written as Equations (4) and (5) (Ho 2006). 
formula
(4)
 
formula
(5)
where qe is the amount of metal ion sorbed onto adsorbent (mmol/g), Ce is the equilibrium concentration of the metal ion in the equilibrium solution (mmol/L). Qm is the monolayer adsorption capacity (mmol/g) and b is the constant related to the free energy of adsorption. Kf is a constant indicative of the relative sorption capacity of the adsorbent (mmol/g) and 1/n is the constant indicative of the intensity of the sorption process. If the value of n is greater than 1, it means a strong interaction between the surface of the adsorbent and adsorbate.

A comparison of all fitted isotherms is shown in Figure S6 (available online), and the isotherm parameters and correlation coefficients are summarized in Table 2. It can be found that the adsorption data for Cs+ and Sr2+ fitted more splendidly to the Langmuir model, with higher correlation coefficients compared to the Freundlich model. Meanwhile, for sorption onto Fe(II)-geopolymer, the Freundlich model was more befitting than the Langmuir model, indicating that the sorption process may be multilayer. In this case, due to n > 1 for sorption onto Fe(II)-geopolymer, it had an increasing tendency for sorption with increasing solid phase concentration. On the other hand, the maximum monolayer sorption capacity values of the adsorbate can be calculated according to the Langmuir model. As for original geopolymer, it had relatively higher maximum adsorption capacity of Cs+ (0.842 mmol/g) than Sr2+ (0.276 mmol/g).

Table 2

Calculated Langmuir, Freundlich and Dubinin-Radushkevich parameters for the sorption of Cs+ and Sr2+ onto original geopolymer, onto Fe(II)-geopolymer

Kinetic modelParameterCs+Sr2+
Langmuir model Qm (mmol/g) 0.842 0.276 0.162 
b (L/mmol) 2.003 11.045 3.694 
RL 0.103 0.050 0.094 
R2 0.977 0.993 0.902 
Freundlich model 2.744 5.752 3.289 
Kf (mmol/g) 0.507 0.253 0.121 
R2 0.965 0.999 0.987 
D-R model β (mol2/kJ20.004 0.002 0.003 
qm (mmol/g) 1.790 0.437 0.329 
E (kJ/mol) 10.937 14.556 12.910 
R2 0.970 0.993 0.930 
Kinetic modelParameterCs+Sr2+
Langmuir model Qm (mmol/g) 0.842 0.276 0.162 
b (L/mmol) 2.003 11.045 3.694 
RL 0.103 0.050 0.094 
R2 0.977 0.993 0.902 
Freundlich model 2.744 5.752 3.289 
Kf (mmol/g) 0.507 0.253 0.121 
R2 0.965 0.999 0.987 
D-R model β (mol2/kJ20.004 0.002 0.003 
qm (mmol/g) 1.790 0.437 0.329 
E (kJ/mol) 10.937 14.556 12.910 
R2 0.970 0.993 0.930 
In addition, one of the essential characteristics of the Langmuir model can be expressed by a dimensionless constant called equilibrium parameters RL. 
formula
(6)
where C0 is the highest initial metal ion concentration (mmol/L). The value of RL indicates the type of isotherm to be irreversible (RL = 0), favorable (0 < RL < 1), linear (RL = 1), or unfavorable (RL > 1). As it can be seen in Table 2, all the RL values for the sorption of Cs+ and Sr2+ onto geopolymer and the sorption of onto Fe(II)-geopolymer were found to be greater than 0 and less than 1, which indicated the favorable sorption isotherms of these ions.
Furthermore, in order to study the nature of these sorption process, the D-R isotherm was also adopted in the form (Misaelides et al. 2018): 
formula
(7)
where qm is the maximum amount of ion that can be sorbed onto unit weight absorbent (mmol/g), and β is the constant related to the sorption energy (mol2/KJ2). ɛ is the Polanyi potential = RT ln(1 + 1/Ce), where R is the gas constant (8.314 J/mol K), and T is the absolute temperature (K).
In addition, the average free energy of sorption could be the free energy change when the ions are transferred to the surface of adsorbent from the solution. The calculation form is expressed in Equation (8). 
formula
(8)
where E is the free energy (kJ/mol). Generally, the free energy in the range of 8–16 kJ/mol indicates the cation exchange process, and a value less than 8 kJ/mol suggests physical adsorption, and one greater than 16 kJ/mol implies chemical adsorption (Misaelides et al. 2018).

The D-R plots of ln qe versus ɛ2 for these sorption processes are given in Figure S6 and the statistical results are shown in Table 2. It can be seen that the sorption processes of Cs+ and Sr2+ onto original geopolymer have the free energies of 9.561 and 14.556 kJ/mol, respectively. This obviously indicated that the sorption processes for Cs+ and Sr2+ are dominantly ion exchange. However, the energy value for the sorption of onto Fe(II)-geopolymer was also lower than 16 kJ/mol, suggesting that the mechanism for the removal of could be not mainly chemical precipitation, and some other interaction should exist.

Sorption mechanisms

Sorption of Cs+ and Sr2+ onto original geopolymer

Compared to other reported adsorbents (Table S1, available online), the geopolymer and its modified type showed high adsorption abilities for Cs+ and Sr2+, and , respectively. Ion exchange with the alkaline ions is considered to be the main reaction in the removal process of Cs+ and Sr2+ by geopolymer, which could be confirmed by the comparisons of released Na+ with the adsorption amount of Cs+ and Sr2+, and SEM-EDX mappings, shown in Figures S7 and S8, respectively. The XRD patterns of original geopolymer, Cs-geopolymer and Sr-geopolymer are presented in Figure S9(a). There was neither a new phase nor obvious change occurred between them. FTIR spectra were collected as shown in Figure S10. (Figures S7–S10 are available online.) Compared with the original geopolymer, there was no significant difference between them. As discussed earlier, the weight loss in the range of 160–330 °C is related to bound water molecules connected to the mobile non-framework cations in the cavities of geopolymer. The less weight loss occurred in Cs-geopolymer (1.21%) than that of original geopolymer (1.62%) in the range of 160–330 °C (Figure 3(a)). However, the weight loss of Sr-geopolymer (1.92%) was higher than that of the original geopolymer. This suggests that the amount of water molecules strongly polarized by Sr2+ is much higher than those polarized by Na+ and Cs+. In this case, hydration numbers of cations should be considered. Specifically, the hydration number (Hn) is the number of those water molecules that undergo strong interaction with ions (Madejová et al. 2002). The hydration numbers for Na+ and Cs+ are significantly smaller (Hn = 3.5 and 2.1, respectively) than for divalent Sr2+ (Hn = 6.4) (Marcus 1994).

Figure 3

(a) TG-DTA curves of original geopolymer, Cs-geopolymer and Sr-geopolymer; (b) FTIR spectra of Fe(II)-modified geopolymer before and after sorption of arsenate; (c) XPS wide scan of Fe(II)-geopolymer before and after sorption of arsenate; (d) Fe 2p peak separation of Fe(II)-geopolymer after sorption of arsenate; (e) As 3d peak separation of Fe(II)-geopolymer after sorption of arsenate.

Figure 3

(a) TG-DTA curves of original geopolymer, Cs-geopolymer and Sr-geopolymer; (b) FTIR spectra of Fe(II)-modified geopolymer before and after sorption of arsenate; (c) XPS wide scan of Fe(II)-geopolymer before and after sorption of arsenate; (d) Fe 2p peak separation of Fe(II)-geopolymer after sorption of arsenate; (e) As 3d peak separation of Fe(II)-geopolymer after sorption of arsenate.

In addition, it is obviously noticed that the original geopolymer has a higher capacity for Cs+ sorption (0.827 mmol/g) than for Sr2+ (0.285 mmol/g). Here, zeolite A, belonging to the framework of cubic space group Fm3c, was adopted as a reference. The sorption amount of Sr2+ onto zeolite A is generally higher than that of Cs+ (Vipin et al. 2016). Furthermore, Cs+ is inclined to go through the window to occupy the position of the eight-member ring since it has a large radius size, while Sr2+ is more likely to replace with the Na+ in the six-member ring (Yoshida et al. 2013). Therefore, there should be mainly two effects including ring size and charge balance controlling their ion exchange processes onto geopolymer. Theoretically, only an eight-member ring or a large one can accommodate Cs+ in the structure, and the six-member ring should be the minimum size for Sr2+. On the other hand, there is only an electrical charge on Cs+, which indicates that Cs+ can be compensated by only one negative charge produced by the AlO4 tetrahedron. However, two negative charges would be needed to make a charge balance for Sr2+. Therefore, the arrangement of SiO4 and AlO4 tetrahedrons in the geopolymer should be considered. Normally, the ratio of Si/Al in zeolite A is 1 and the arrangement of SiO4 and AlO4 tetrahedrons should be alternated (–Si–O–Al–O–Si–O–Al–). If the distance becomes larger, it would be difficult to make a charge balance between Sr2+ and the AlO4 tetrahedron. Theoretically, the Si/Al ratio in the synthesized geopolymer is approximately 1.89. Therefore, there are at least three kinds of arrangements of SiO4 and AlO4 in the geopolymer: –Al–O–Si–O–Al–, –Al–Si–O–Si–O–Al– and –Al–Si–O–Si–O–Si–O–Al–. Only the first can provide a proper ion exchange site for Sr2+. It is extremely difficult to quantify their amount in the geopolymer structure because of its amorphous state. But it is confirmable that the arrangement of –Al–O–Si–Al– should be limited. This should be the main reason why the sorption amount of Sr2+ is relatively lower than that of Cs+.

Sorption of arsenate onto Fe(II)-geopolymer

In the present work, the geopolymer was modified and converted into Fe(II) form by a cation exchange process and further applied for removal. Figure S9(b) presents the XRD patterns of Fe(II) geopolymer before and after sorption. There was no new phase appeared on the XRD pattern of Fe(II) geopolymer after sorption. FTIR spectra of Fe(II) geopolymer before and after sorption are given in Figure 3(b). It can be found that there are two small peaks in the range of 750 to 850 cm−1 on the spectrum of Fe(II)-geopolymer after sorption, which could be assigned to the vibrations mode of As-O in arsenate (Sun & Doner 1996). Furthermore, another two peaks occurred at 400–600 cm−1, representing the bond of Fe-O, and this indicated that arsenate removal might be due to the co-precipitation and/or chemisorption of arsenate on iron hydroxide. The spectra of XPS wide scans of Fe(II)-geopolymer before and after sorption are shown in Figure 3(c). Small peaks representing As3d and As3p signals also occurred, which indicated that had been immobilized on the surface of Fe(II)-geopolymer. It should be convincing that ferrous ion played the key role in the sorption of onto Fe(II) geopolymer. Figure S11 presents SEM-EDX elemental mapping. The distribution of As showed a good correspondence with the Fe, and this further certified the above deduction. On the other hand, the XPS spectra of Fe 2p peaks for Fe(II) geopolymer before and after sorption are presented in Figure S12. (Figures S11 and S12 are available online.) Compared to the Fe 2p peaks of Fe(II) geopolymer before sorption of , the two peaks for the adsorbent after sorption exhibited a small shift to higher binding energy. This might imply that Fe(II) might be partially oxidized into Fe(III) during sorption of arsenate. The peak fitting results of Fe and As in Fe(II) geopolymer after sorption are shown in Figure 3(d) and 3(e). The narrow scan for the As 3d spectrum showed characteristic peaks located at 45.7 and 46.4 eV, which were attributed to EB [As 3d] for (Muthu Prabhu et al. 2018). In addition, the binding energies for Fe2+ and Fe3+ could be in the range of 709.9–715 and 711.1–719.8 eV (Simsek & Beker 2014). From Figure 3(d), it can be known that a part of Fe(II) has been oxidized into Fe(III), and the molar ratio of Fe(III)/As on the surface of the adsorbent could be calculated using sensitivity factors (Wagner 1983), showing as 0.64. It can be found in Figure S13 (available online) that the concentrations of dissolved Fe and As adsorption amount showed a reverse trend, which suggested that the removal process of should be chemical sorption. The chemical formation of FeAsO4 (Ksp ≈ 5.8 × 10–21 at 25 °C) should be the mainly expected removal mechanism. However, from the calculation results from the D-R isotherm, the free energy (12.91 kJ/mol) was lower than 16 kJ/mol, and the calculated value of Fe/As on the surface of the Fe(II) geopolymer after sorption is obviously lower than the Fe/As ratio in the chemical formula of FeAsO4. This indicates that there might be another sorption mode happening. Specifically, should be a dominant species in a solution with a pH of 5 (Nicomel et al. 2016). After oxidation of Fe2+ into Fe3+, some Fe3+ might present the formulas of (x = 1 or 2) (Nekhunguni et al. 2017). Thus, sorption might have partially occurred through the formation of inner sphere complexes via ion exchange reaction, which could be predominantly attached by bidentate linkages (Velazquez-Peña et al. 2019), as Equation (9) indicates. Thus, precipitation and complexation could be the main mechanisms for removal using Fe(II)-modified geopolymer. 
formula
(9)

CONCLUSIONS

Coal fly ash-based geopolymer and its derivative (Fe(II)-modified geopolymer) were successfully synthesized for removals of Cs+ and Sr2+, and , respectively, from aqueous solutions. The results illustrated that the original geopolymer had a much larger surface area (131.4 m2/g) than coal fly ash (1.1 m2/g). According to the kinetic fitting results, each sorption process fitted better with the pseudo-second-order model and the overall rate constants appeared to be controlled by chemical sorption. On the other hand, the sorption processes of Cs+ and Sr2+ onto the original geopolymer have better agreement with the Langmuir model. However, the Freundlich model is more befitting for sorption onto Fe(II) geopolymer. Based on the D-R simulation results, ion exchange is believed to be the main mechanism for sorption of Cs+ and Sr2+ onto the original geopolymer. Specifically, for Cs+, ring size is the main factor controlling the ion exchange process. However, both ring size and arrangement of SiO4 and AlO4 tetrahedrons have the most significant effects on the ion exchange process for Sr2+. This has a close relationship with the molar ratio of Si/Al in the geopolymer structure. On the other hand, for the sorption of onto Fe(II) geopolymer, chemical formation of FeAsO4 only partially contributes to the removal of from solution. In addition, sorption might have occurred through the formation of inner-sphere complexes via ion exchange reaction, which could be predominantly attached by bidentate linkages.

ACKNOWLEDGEMENTS

This research was supported to KS by the JSPS (Japan Society for the Promotion of Science) Kaken Kiban A project (No. JP16H02435) and to TQ by the China Scholarship Council (No. 201706420068).

REFERENCES

REFERENCES
Al-Mashqbeh
A.
,
Abuali
S.
,
El-Eswed
B.
&
Khalili
F. I.
2018
Immobilization of toxic inorganic anions (, and in metakaolin based geopolymers: a preliminary study
.
Ceramics International
44
,
5613
5620
.
Benaliouche
F.
,
Hidous
N.
,
Guerza
M.
,
Zouad
Y.
&
Boucheffa
Y.
2015
Characterization and water adsorption properties of Ag-and Zn-exchanged A zeolites
.
Microporous and Mesoporous Materials
209
,
184
188
.
Cho
Y.-K.
,
Yoo
S.-W.
,
Jung
S.-H.
,
Lee
K.-M.
&
Kwon
S.-J.
2017
Effect of Na2O content, SiO2/Na2O molar ratio, and curing conditions on the compressive strength of FA-based geopolymer
.
Construction and Building Materials
145
,
253
260
.
El-Eswed
B. I.
2018
Solidification Versus Adsorption for Immobilization of Pollutants in Geopolymeric Materials: A Review
. In:
Solidification
.
InTech
Open, London, UK
, pp.
78
110
.
Ho
Y.-S.
2006
Isotherms for the sorption of lead onto peat: comparison of linear and non-linear methods
.
Polish Journal of Environmental Studies
15
,
81
86
.
Lopes
J. H.
,
Nogueira
F. G.
,
Gonçalves
M.
&
Oliveira
L. C.
2015
Modified zeolite with transition metals Cu and Fe for removal of methylene blue from aqueous medium: mass spectrometry study
.
Bulletin of Chemical Reaction Engineering & Catalysis
10
,
237
248
.
Madejová
J.
,
Janek
M.
,
Komadel
P.
,
Herbert
H.-J.
&
Moog
H.
2002
FTIR analyses of water in MX-80 bentonite compacted from high salinary salt solution systems
.
Applied Clay Science
20
,
255
271
.
Misaelides
P.
,
Sarri
S.
,
Kantiranis
N.
,
Noli
F.
,
Filippidis
A.
,
de Blochouse
B.
,
Maes
A.
&
Breynaert
E.
2018
Investigation of chabazitic materials as Cs-137 sorbents from cementitious aqueous solutions
.
Microporous and Mesoporous Materials
266
,
183
188
.
Nekhunguni
P. M.
,
Tavengwa
N. T.
&
Tutu
H.
2017
Investigation of As(V) removal from acid mine drainage by iron (hydr)oxide modified zeolite
.
Journal of Environmental Management
197
,
550
558
.
Nicomel
N.
,
Leus
K.
,
Folens
K.
,
Van Der Voort
P.
&
Du Laing
G.
2016
Technologies for arsenic removal from water: current status and future perspectives
.
International Journal of Environmental Research and Public Health
13
,
62
.
Rożek
P.
,
Król
M.
&
Mozgawa
W.
2018
Spectroscopic studies of fly ash-based geopolymers
.
Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy
198
,
283
289
.
Runtti
H.
,
Luukkonen
T.
,
Niskanen
M.
,
Tuomikoski
S.
,
Kangas
T.
,
Tynjälä
P.
,
Tolonen
E.-T.
,
Sarkkinen
M.
,
Kemppainen
K.
,
Rämö
J.
&
Lassi
U.
2016
Sulphate removal over barium-modified blast-furnace-slag geopolymer
.
Journal of Hazardous Materials
317
,
373
384
.
Velazquez-Peña
G. C.
,
Solache-Ríos
M.
,
Olguin
M.
&
Fall
C.
2019
As (V) sorption by different natural zeolite frameworks modified with Fe, Zr and FeZr
.
Microporous and Mesoporous Materials
273
,
133
141
.
Vipin
A. K.
,
Ling
S.
&
Fugetsu
B.
2016
Removal of Cs+ and Sr2+ from water using MWCNT reinforced Zeolite-A beads
.
Microporous and Mesoporous Materials
224
,
84
88
.
Wagner
C.
1983
Sensitivity factors for XPS analysis of surface atoms
.
Journal of Electron Spectroscopy and Related Phenomena
32
,
99
102
.
Yoshida
K.
,
Toyoura
K.
,
Matsunaga
K.
,
Nakahira
A.
,
Kurata
H.
,
Ikuhara
Y. H.
&
Sasaki
Y.
2013
Atomic sites and stability of Cs+ captured within zeolitic nanocavities
.
Scientific Reports
3
,
2457
.

Supplementary data