3.1. Physio-Chemical Properties of Materials
Morphological characteristics were initially used to investigate the physical properties of materials. The morphology of prepared materials is illustrated in
Figure 1. Similar to other studies [
28], PA-1 (washed with deionized water) had rough surfaces. However, the particle morphology of HCl modification (PA-2) was different from PA-1, leading to an irregular surface and porosity. The shape and appearance showed a rough surface and higher porosity than PA-1 due to HCl modification, which could be used to remove surface impurities and enhance the physical properties of materials. Therefore, fluoride adsorption into the pore spaces increased after the materials were acid-modified [
34].
The quantitative chemical compositions of PA-1 and PA-2 were analyzed using an energy-dispersive X-ray microanalyzer (SEM-EDS). The chemical composition of PA-1 exhibited the types of elements shown in
Table 2. The highest proportion (% wt) of PA-1 was Mn, followed by Fe, O, C, N, Al, and Si, respectively. The elements in the prepared samples showed a similar composition to that of the original pyrolusite, which contained the complex of MnO
2, CaO, Fe
2O
3, SiO
2, Al
2O
3, and MgO [
27]. In addition, the PA-2 (washed with 0.1 N HCl) showed a similar trend to the PA-1, however, some slight changes occurred. The Si and N disappeared, and some mineral content such as Al was reduced after acidic modification [
35].
Figure S1 in the supplementary materials shows the results of surface functional group characterization using FTIR of the original PA (PA-1) and acid-modified-PA (PA-2). FITR analysis was investigated in a range of 400–4000 cm
−1. A pyrolusite peak (MnO
2) was exhibited in both PA-1 and PA-2 wavenumber brands at 650–800 cm
−1 [
32]. The surface changes of acid-modified-PA (PA-2) were observed as a new brand between 1700–1800 cm
−1 and 2900 cm
−1, indicating that the C=O and C-H stretching vibrations of hydroxyl groups are associated with hydrochloric acid molecules via hydrogen bonding [
36,
37]. As a result, the success of acidic modification could be demonstrated.
The porosity and surface area of both PAs were investigated by the nitrogen adsorption isotherm (
Figure 2 and
Table 3) via the BET and BJH methods, respectively. The calculated BET surface areas of PA-1 (washed with DI water) and PA-2 (washed with acid) were 111.48 and 108.93 m
2/g, respectively. The pore sizes of PA-1 and PA-2 materials, measured by Barret–Joyner–Halenda (BJH) methods, were 3.65 and 4.62 nm, respectively. PA-2 had larger pores than PA-1 because it had been activated with an acid solution to improve its porous properties. In addition, the pore volumes of PA-1 and PA-2 were 0.112 and 0.107 cc/g, respectively. The pore volume of both adsorbents showed a small reduction in PA-2 compared with PA-1, which indicated the possession of PA-1 by residual HCl molecules. Furthermore, the hysteresis loops of PA-1 and PA-2 represent mesopore adsorbents with pore sizes ranging from 2 to 50 nm. The zeta potentials of PA-1 and PA-2 at varying ranges of pH are shown in
Figure S2 in the supplementary materials. At the point of zero charge (pH
pzc), the pH of both materials was approximately 4.00.
3.2. Adsorption Kinetics Study
The adsorption kinetics studied of fluoride on PA-1 and PA-2 materials were investigated to determine the rate of adsorption, as displayed in
Figure 3. The adsorption rate reached equilibrium in all adsorbents after 60 min, indicating the potential for rapid fluoride removal due to the rapid rate at which an equilibrium state was reached between the adsorbate and the adsorbent, and the result followed a similar trend seen with other ore adsorbents [
24]. An important and practical component of this rapid adsorption is the use of tiny reactor volumes, which guarantee great efficiency at a low cost [
38]. Comparing PA-1 and PA-2, the adsorption capacity of PA-2 was higher than PA-1. This may be possible with a surface modification of 0.1 N HCl to enhance positive surface charge [
35,
39]. Although the BET results for PA-1, including surface area and pore volume, are higher than those for PA-2, the adsorption capacity for PA-1 is lower. This could confirm that the adsorption mechanism of pyrolusite ore on fluoride is not only a physical adsorption mechanism. Another result of the material characterization could confirm the adsorption mechanism and that is the pH
pzc of the materials. The point of zero charge for both adsorbents was 4.00; however, the pH of the solution was greater than 4.00, indicating that the adsorbent surface was negatively charged (pH > pH
pzc), resulting in low adsorption capacities in both adsorbents [
38]. This would confirm that the adsorption of prepared ore (PA) would make it possible to allow for a variety of types of adsorption mechanisms, such as physical, chemical, and the effect of electrostatic interaction between the material and the solution. If the surface of the material is modified to be positively charged, it will be suitable for use with a fluoride solution (a negatively charged solution) to enhance its adsorption capacity [
34].
Two different kinetic models including pseudo-first order and pseudo-second order equations were applied to quantitatively determine the rate of defluorination. The equation of pseudo-first order and pseudo-second order models are given by Equations (4)–(6) [
21,
40,
41], respectively.
The pseudo-first order equation is presented by the following equation:
Which can be rearranged in linear form as follows:
where
qt is the adsorption capacity at time
t (mg/g),
qe is the adsorption capacity at equilibrium (mg/g),
t is time (min), and
kp1 is the pseudo-first-order rate constant (min
−1).
The pseudo-second order equation is applied to explain the chemical reactions of heterogeneous materials. This model is presented by the following equation:
where
qt is the adsorption capacity at time
t (mg/g),
qe is the adsorption capacity at equilibrium (mg/g),
t is time (min), and
kp2 is the pseudo-second order rate constant (g/mg∙min).
Based on the pseudo-second order equation, the initial rate of adsorption, h (mg/g/min) at
t = 0 can be calculated using Equation (7)
All calculations of the adsorption kinetic models are summarized in
Table 4. The kinetic data of PA-1 and PA-2 represented a good correlation (R
2) with the pseudo-second order model (R
2 = 0.9952 and 0.9957, respectively.). However, the pseudo-first order model was not well fitted with both materials (R
2 less than 0.7). In addition, the adsorption capacity from the pseudo-second order model (q
e,
cal) was close to the experimental results (q
e,
exp). The results indicated that a good performance of both PAs and fluoride probably represented a combination of chemical sorption and a dependency on the effect of the active site of adsorbent and adsorbate, especially in the presence of surface modification with an acid solution [
19]. When considering the K
2 parameter from the pseudo-second order model, the rate of PA-2 was higher than that of PA-2 due to the effect of a flat surface resulting in an increased speed of fluoride adsorption [
28].
3.3. Adsorption Isotherms
The adsorption isotherm of two PA materials (PA-1 and PA-2) on fluoride removal are shown in
Figure 4. Adsorption studies of fluoride on PA-1 and PA-2 were conducted at pH 7 controlled by a phosphate buffer. The graphical plot of all obtained isotherm data exhibited a linear function of fluoride adsorption. From the results, the adsorption isotherm of PA-2 showed higher adsorption capacities due to higher pore volume and perhaps because of acid modification slightly increasing the positive charged and active site properties [
42].
Four isotherm models (linear, Langmuir and Freundlich and Sips model) were used to investigate the adsorption behavior of fluoride on pyrolusite ore adsorbents (
Table 5). [
41,
43]. The concentration range of 2–15 mg/L was used due to their occurrence with a solid to liquid ratio of 15 g/L.
The linear model equation is shown as Equation (8)
where
is the linear partition and
is the concentration at equilibrium of fluoride.
The Langmuir isotherm defines monolayer sorption onto the surface, with sorption locating only at some sites. There are no interactions between the molecules. The model equation is shown as Equations (9) and (10):
Which can be rearranged in the linear form as follows:
where
q0 is the amount of fluoride adsorbed per unit weight of PA in forming a complete monolayer on the surface (mg/g),
qe is the total amount of fluoride adsorbed per unit weight of PA at equilibrium (mg/g),
Ce is the concentration of the fluoride in the solution at equilibrium (mg/L), and
KF is the constant related to the energy of sorption (L/mg).
The Freundlich isotherm can be applied with non-ideal sorption on heterogeneous surfaces and with multilayer sorption. The model is shown by the following Equations (11) and (12):
Which can be rearranged in the linear form as follows:
where
qe is the total amount of fluoride adsorbed per unit weight of PA at equilibrium (mg/g),
Ce is the concentration of the fluoride in the solution at equilibrium (mg/L),
KF is the Freundlich constant which can express the capacity of the adsorption process (L/g), and
n is the Freundlich constant, which explains the concentration of adsorption (dimensionless).
The Sips isotherm model is a combination of the Langmuir and Freundlich isotherm models, which are appropriate for predicting adsorption on heterogeneous surfaces. The equation is shown as Equation (13):
where
is the Sips isotherm exponent,
KS (L/µg) is the Sips adsorption constant and
is the maximum adsorption capacity (µg/g).
Based on isotherm results, the adsorption mechanism could be explained by the mentioned isotherm model of fluoride adsorption by PA-1 and PA-2. The correlation rankings, in ascending order (R
2 highest to lowest), of both PA-1 and PA-2 were Freundlich > Sips
Linear > Langmuir. This could be an indication of monolayer adsorption on the surface of PAs materials [
44]. The statement of the Freundlich isotherm model defines both the heterogeneity of the surface and the exponential distribution of the active sites. To confirm the performance results of adsorption, the value of 1/
n from the fitted isotherm model was used for prediction. When 1/
n is higher than zero and does not exceed 1, the adsorption process is favorable. As a result, this study found 1/
n values of 0.51 and 0.56 for PA-1 and PA-2, respectively, confirming that the fluoride adsorption process on PAs performed well, particularly at acid-modification (PA-2) sites [
44].
In addition, the research that has been done on the adsorption of fluoride by low-grade pyrolusite is limited and needs to be supplemented with additional research for further investigation. Even though neither the original nor the acid-modified ore had an especially high adsorption capacity, both types of ore are still suitable for use as non-commercial materials in the process of reducing the amount of fluoride that is present in groundwater in order to comply with the regulatory standard that is in place. When the adsorption capacity of prepared pyrolusite in this study was compared with that of low-cost or abundant materials on fluoride removal, the results of the comparison were shown in
Table 6. These materials included iron ore, dolomite ore, and
Polygonum orientale Linn. The adsorption capacity of all materials on fluoride adsorption is based on the Freundlich adsorption capacity. When comparing the same initial fluoride concentration (10 mg/L) with abundance ore, the results demonstrated that PA-1 and PA-2 have a higher fluoride adsorption capacity than that of dolomite ore; this may be due to the smaller pore volume of dolomite (0.0031 cc/g) than that of either PA-1 or PA-2, which were 0.112 and 0.107 cc/g, respectively. Although the adsorption capacity of iron ore showed the highest efficiency, the equilibrium reaction time was 120 min at pH 6, which was higher than PA-1 and PA-2’s (60 min) at pH 7 because of the difference in temperature and pH in the reaction. Furthermore, because the selected materials are abundant, the cost of all adsorbents is only in the surface modification of the materials, such as acid or metal oxide modification. When comparing the modification cost, for example, of
Polygonum orientale Linn., modified with Al
2(SO
4)
3 (PO-Al) and acid-modified pyrolusite (PA-2), the functionalization cost of PO-Al will be slightly higher than PA-2 due to the price of Al
2(SO
4)
3 being higher than hydrochloric acid.
According to the findings, this sort of adsorbent allows the prepared pyrolusite ore, as well as acid-modified materials, to be compatible with other types of materials. Even though some materials, such as iron ore, demonstrated good performance [
24], pyrolusite is an abundant ore option from the northern region of Thailand that will be able to be used at real and practical sites for fluoride removal, especially in places that contain a slightly higher fluoride level.