Excess fluoride in drinking water is harmful to human health and therefore it needs to be removed from water before consumption. The potential of locally available mixed Mukondeni clay soils (MMCS) as a cheap adsorbent for the removal of fluoride from aqueous solution was investigated. Characterization of MMCS was done by X-ray fluorescence, X-ray diffraction, scanning electron microscopy, Fourier transform infrared and Brunauer Emmett Teller. Cation exchange capacity and point of zero charge of the clays were determined using standard methods. Parameters optimized included: contact time, adsorbent dosage, initial fluoride concentration, pH and temperature. Optimization experiments were done in batch procedures. The results showed that the optimum conditions for the defluoridation of water using MMCS are 60 min, 1.5 g, 9 mg/L, 1.5/100 S/L ratios, pH of 2 and a temperature of 25 °C. The equilibrium isotherm regression parameter (R2 = 0.95) showed that the Freundlich isotherm gave a better fit than the Langmuir isotherm (R2 = 0.52) which indicates multilayer adsorption. Kinetic studies revealed that the adsorption followed pseudo second order kinetics. This study indicated that locally available MMCS are good in the defluoridation of groundwater but modification through blending with metal oxide modified clays can enhance their adsorption capacity.

INTRODUCTION

Freshwater is critical to life on earth and in some circumstances it can be the medium for life itself. However, reservoirs have water that is usually altered from its original mineral concentration by interactions with mineral weathering (Banerjee 2015). This alteration especially occurs in groundwater and one common example of groundwater contamination through water–rock interaction is leaching of fluoride from fluoride bearing rocks.

The World Health Organization (WHO) classifies fluoride as one of the contaminants of water for human consumption. WHO has set a guiding value of 1.5 mg/L for fluoride in drinking water (WHO 2011). Studies done in South Africa have revealed that the occurrence of dental fluorosis in many South Africans is related to the fluoride content of groundwater used for drinking purposes (WRC 2001). A small village in the Limpopo province called Siloam sources water from boreholes with a fluoride content above 5 mg/L (Gitari et al. 2013). Therefore, mitigation of fluorosis is a very significant area of research and needs continued attention until a reliable defluoridation technology is developed.

Several defluoridation methods including precipitation, ion exchange and membrane processes have been developed for defluoridation purposes (Vázquez-Guerreroa et al. 2016). However, the weaknesses of a lot of these techniques are numerous. They include high operational and maintenance costs, secondary pollution, production of toxic sludge and complex processes involved in the treatment (Bhatnagar et al. 2011). In light of the above, consideration must be given to those methods that reduce the capital and running costs. Moreover they have to be affordable and available to small communities, especially in rural areas located in developing countries that do not have adequate and effective water treatment technologies. Among the above mentioned methods, adsorption is regarded as the most promising method for the defluoridation of water due to the ease of operation, lower cost and being a relatively environmentally friendly process (Thakre et al. 2011).

A number of studies have been performed using raw and modified clays for the defluoridation of contaminated water (Masindi et al. 2014; Vinati et al. 2015; Zhang et al. 2016). The factors found to influence fluoride sorption include solution pH, clay surface area and the presence of exchangeable cations capable of forming fluoride precipitates. In the Limpopo province of South Africa, Mukondeni clay soils have been used for pottery making and ceramic water filters for a very long time but their physicochemical characteristics had never been studied to determine how suitable they are for these purposes. A variety of Mukondeni clay soils have recently been getting more attention in water treatment applications (Denga 2013; Abebe et al. 2014). Thus, it becomes significant that the basic characteristics of these soils are evaluated and their scientific application in the treatment of water be assessed. Moreover these clay soils are locally and readily available, hence a breakthrough in their ability to remove contaminants from drinking water will come as an advantage to the local community, most of whom are poor and cannot afford expensive water treatment methods. This study was undertaken to examine for the first time the physicochemical and mineralogical characterization of mixed Mukondeni clay soils (MMCS). Furthermore to evaluate MMCS's fluoride adsorption capacity, isotherms, kinetics and mechanisms of fluoride adsorption.

Sampling and adsorbent preparation

All chemicals used in the present study were of analytical reagent grade. MMCS were collected from Mukondeni Village in Limpopo province. Fluoride rich water was collected from a borehole in Siloam village in Limpopo province. Mukondeni clay soils were washed by rinsing with MilliQ water. The residual clay was left to dry in the oven for 12 h at 105 °C. After drying the samples were milled into fine powder and passed through a 250 μm sieve.

Characterization of the adsorbent

Mineralogical composition of the clay was determined using a PANalytical X'Pert Pro powder diffractometer in θθ configuration with an X'Celerator detector and variable divergence- and fixed receiving slits with Fe filtered Co-Kα radiation (λ = 1.789 Å). Elemental composition was determined using X-ray fluorescence (XRF) (Thermo Fisher ARL-9400XP + Sequential XRF equipped with WinXRF software). Morphology was determined using a Hitachi X-650 Scanning Electron Microanalyser equipped with a CDU-lead detector at 25 kV. Functional groups were determined using a Fourier transform infrared (FTIR) flow-through continuous reactor (Bruker Alpha spectrometer). Surface area was determined using Brunauer Emmett Teller (BET) analysis (Micromeritics Tristar II, Norcross, GA, USA). The cation-exchange capacity of the clay was determined using the ammonium acetate method (Chapman 1965). Point of zero charge (PZC) was determined by solid addition method as described by Izuagie et al. (2016).

Batch adsorption experiments and fluoride analysis

A stock solution of 1,000 mL of 1,000 mg/L fluoride was prepared by dissolving 2.210 g of NaF in Milli-Q (18.2 MΩ/cm) water in a litre volumetric flask. Lower concentrations of fluoride were prepared from the stock solution by serial dilution. 100 mL of the synthetic fluoride solution was pipetted into a 250 mL plastic bottle and a desired amount of the adsorbent was added to it and then shaken at different contact times (1–180 min, 1, 2 and 3 g, 10 mg/L F, 25 °C room temperature, 100 mL, 250 rpm, pH = 6.423) on a reciprocating shaker in order to attain equilibrium. The bottles were removed from the shaker and then allowed to stand for 5 min. The solution was then filtered using 0.45 μm pore cellulose nitrate membranes and the filtrate was analysed for residual fluoride concentration. Fluoride concentration was measured using an ion selective electrode (ISE) Thermo Scientific Orion Star A215 pH/ Conductivity Benchtop Meter (USA) attached with 8157BNUMD Orion ROSS Ultra Triode pH/ATC electrode. A similar ion meter coupled with pH electrode was used for measuring the pH of the treated samples. The total ionic strength adjusting buffer (TISAB III) was added in 1:10 (mL) ratio to the sample and standard solutions in order to regulate the ionic strength and to maintain pH at between 5.2 and 5.5 optimum for the F selective electrode. For quality control and assurance, inter laboratory analysis was done in which fluoride measurements were taken by both ion chromatography (Metrohm 850 Professional IC) and ISE to check if the results were in the same range using different instruments of analysis, but ISE results were reported for fluoride measurements. The same procedure was repeated for different dosages (0.1–2 g, 60 min contact time, 10 mg/L F, 25 °C room temperature, 250 rpm, pH = 5.937); initial fluoride concentrations (1–15 mg/L, 60 min contact time, 1 and 2 g dosage, 25 °C room temperature, 100 mL, 250 rpm, pH = 6.111); temperature (298, 308 and 318 K, 60 min contact time, 1 and 2 g dosage, 100 mL, 250 rpm, pH = 6.111) and pH (2–12, 60 min contact time, 1.5 g, dosage, 9 mg/L F, 25 °C room temperature, 100 mL, 250 rpm). The fluoride removal ability of MMCS was then tested on field water. All the experiments were carried out at room temperature (25 ± 3 °C). The effect of pH on fluoride removal was studied by adjusting the pH of the solution using 0.1 mol/L HCl and 0.1 mol/L NaOH solutions. To validate the reproducibility of the results, the experiments were carried out in triplicate and the mean values were reported. Blank experiments showed no detectable adsorbed fluoride ions by the filtration of the supernatants. Equations (1) and (2) were used to calculate the percentage of removal and adsorption capacity respectively. 
formula
1
where Co is initial fluoride ion concentration and Ce is equilibrium fluoride ion concentration in mg/L. 
formula
2
where Co is initial F concentration (mg/L), Ce is F concentrations at equilibrium (mg/L), V is volume of solution (L) and m is weight of the adsorbent (g).

Regeneration studies

The potential for reusing fluoride adsorbed MMCS was tested by adding 2 g in 100 mL of a fluoride solution (9 mg/L) for adsorption at 25 °C ± 1 °C and a shaking speed of 250 rpm for 1 h. The solution was then filtered. The fluoride-loaded adsorbent was shaken in 250 mL of 0.01 mol/L NaOH and in 250 mL distilled water afterwards. Four adsorption–desorption cycles were carried out in the same experimental conditions and the amount of fluoride adsorbed was determined in each cycle.

Chemical stability of the adsorbent

The chemical stability of MMCS was investigated by measuring the concentration of dissolved metals in the solution after adsorption. Two grams of MMCS were added to 100 mL of a 9 mg/L fluoride solution over a pH range of 2–12 and defluoridation experiments were carried out. The supernatant was separated through filtration using a 0.45 μm pore cellulose nitrate membranes after shaking for 60 min at 250 rpm using the Stuart reciprocating shaker. The concentration of dissolved metals was measured by inductively coupled plasma mass spectrometry (ICP-MS, DIONEX ICS-2100 from Thermo).

RESULTS AND DISCUSSION

Physicochemical and mineralogical characterization of MMCS

Cation exchange capacity analysis

Table 1 shows the concentration of exchangeable cations in MMCS. The results showed a high concentration of exchangeable Na+ at both pHs and Ca2+ had the least concentration. Since cation exchange capacity (CEC) is an intrinsic property of soil defining the concentration of negatively charged sites on soil colloids that can adsorb exchangeable cations, the identified exchangeable cations in MMCS show that it can be a good material for use in the adsorption process.

Table 1

CEC of MMCS

 Exchangeable cations
CEC (meq/100 g)
pHNa+K+Ca2+Mg2+Total
5.4 44.79 1.4 0.40 24.00 70.59 
7.4 111.30 0.40 24.00 137.70 
 Exchangeable cations
CEC (meq/100 g)
pHNa+K+Ca2+Mg2+Total
5.4 44.79 1.4 0.40 24.00 70.59 
7.4 111.30 0.40 24.00 137.70 

PZC analysis

The pH where the net total particle charge is zero is called the PZC. If the pH of a soil is above its PZC, the soil surface will have a net negative charge and predominantly exhibit an ability to exchange cations, while the soil will mainly retain anions if its pH is below its PZC (Appela et al. 2003). Figure 1 shows that MMCS have a PZC at pH 6.5 which is slightly acidic. This means that adsorption of F using Mukondeni clay soils will be effective in acidic conditions below pH 6.5 because the surfaces will be positively charged, enhancing the uptake of fluoride.
Figure 1

PZC of MMCS.

Figure 1

PZC of MMCS.

X-ray diffraction analysis

Characterization of the structures of MMCS was carried out by X-ray diffraction (XRD) as shown in Figure 2. The diffraction pattern was found to be typical for layered structured clays (Gitari et al. 2013; Masindi et al. 2014). The diffraction pattern exhibited a crystal line structure with an intense reflection at about 20 ° and 27 ° (2θ) characteristic of quartz and four other less intense peaks occurring between 42 ° and 68 ° (2θ). Quartz is made up of a continuous framework of SiO4, suggesting that MMCS are silicate materials. The spectrum also shows the presence of vermiculite and albite which are Mg3Si4O10(OH)2 and Na(AlSi3O8) respectively.
Figure 2

XRD diffractogram of MMCS.

Figure 2

XRD diffractogram of MMCS.

XRF analysis

Table 2 presents the major chemical constituents of MMCS. The high concentration of SiO2 and Al2O3 confirms that this clay is an alumino-silicate. There are however other metal oxides like Fe2O3, MgO and Na2O. These metal oxides contribute towards the fluoride adsorption and hence their presence in these clay soils plays a significant role in the adsorption process.

Table 2

Elemental composition of MMCS

Element as oxides% w/w
Al2O3 14.36 
CaO 1.75 
Cr2O3 0.04 
Fe2O3 5.52 
K20.9 
MgO 2.68 
MnO 0.08 
Na22.53 
P2O5 0.04 
SiO2 61.76 
TiO2 0.57 
L.O.I. 9.3 
Total 99.53 
Element as oxides% w/w
Al2O3 14.36 
CaO 1.75 
Cr2O3 0.04 
Fe2O3 5.52 
K20.9 
MgO 2.68 
MnO 0.08 
Na22.53 
P2O5 0.04 
SiO2 61.76 
TiO2 0.57 
L.O.I. 9.3 
Total 99.53 

Scanning electron microscopy analysis

Scanning electron microscopy (SEM) images of MMCS before and after defluoridation are presented in Figure 3(a) and 3(b) respectively. Figure 3(a) reveals a rough texture with an irregular structure and shapes. However after defluoridation, as shown in Figure 3(b), the clay soils appear to have small particle sizes with undefined shapes. The adsorption of fluoride made the clay soils show agglomerates of very fine particles clustered together with some appearing threadlike and others having very flat surfaces.
Figure 3

SEM images of MMCS before (a) and after (b) defluoridation.

Figure 3

SEM images of MMCS before (a) and after (b) defluoridation.

BET analysis

Results in Table 3 show that MMCS have a BET surface area of 35.4613 m2/g. The surface area is higher compared to typical soils but lower than pure phase goethite and illite, which range from 45 to 169 and 65 to 100 m2/g, respectively (Langmuir 1997). Compared to other adsorbents used in defluoridation, it has a moderate surface area. A study by Gitari et al. (2013) showed that bentonite clay has a surface area of 16.0151 m2/g while Goswami & Purkait (2013) used schwertmannite with a surface area of 314.5532 m2/g.

Table 3

Surface area of Mukondeni clay soils

ParameterMukondeni clay soils
Surface area (m²/g)   
Single point surface area 35.4612 
BET surface area 43.2077 
ParameterMukondeni clay soils
Surface area (m²/g)   
Single point surface area 35.4612 
BET surface area 43.2077 

FTIR analysis

FTIR spectroscopy was utilized to indicate the functional groups present on the surface MMCS. The presence of absorption bands at 457, 529 and 670 cm−1 on the spectrum shown in Figure 4 corresponds to the Si-O-Si bending hence showing that that the material understudy is a silicate (Madejova and Komadel, 2001). This corroborates the XRD findings showing that MMCS are silicate materials. A very high absorbance at wavelength number 900 and 995 cm−1 which corresponds to OH deformations linked to Al3+ was observed, hence showing that aluminium is a building block of the clay minerals in the soils. The slight difference of band intensity at around 900 cm−1 (consisting of OH groups) indicates fluoride sorption by the clay. Since OH and fluoride ions have very similar dimensions, they can replace each other in such a way that Al-F complexes can be formed by the interaction of fluoride and hydroxide ions with Al3+ without breaking the bridging Al-O-Al bonds (Vázquez-Guerreroa et al. 2016).
Figure 4

FTIR spectra of MMCS before and after defluoridation.

Figure 4

FTIR spectra of MMCS before and after defluoridation.

Batch adsorption experiments

Effect of contact time

In order to establish the equilibrium time of fluoride adsorption onto MMCS, the effect of contact time was evaluated. Figure 5(a) shows that there was very little fluoride removal in the first 15 min and a rapid rise at 30 min followed by a decrease again. The increase in fluoride percentage removal with an increase in contact time could be due to the fact that as time increased, the fluoride ions were finding the unoccupied adsorption sites and getting adsorbed but with increase in time, the adsorption sites were exhausted and fluoride removal was reduced. A similar change in the removal efficiency with bentonite was attributed to progressive loss of solute concentration gradient as the initially vacant adsorbent sites were occupied (Guo & Reardon 2012). Similar observations were reported by Ismail & AbdelKareem (2015) and Thakre et al. (2011). It can also be seen that after 60 min no significant fluoride removal was observed indicating that the reaction had reached equilibrium.
Figure 5

(a) Variation of % F removal with contact time (1, 2 and 3 g, 10 mg/L F, 25 °C room temperature, 100 mL, 250 rpm, pH = 6.423). (b) Variation of % F removal with adsorbent dosage, (60 min contact time, 10 mg/L F, 25 °C room temperature, 250 rpm, pH = 5.937). (c) Variation of % F removal with initial concentration (60 min contact time, 2 g dosage, 25 °C room temperature, 100 mL, 250 rpm, pH = 6.111). (d) Variation of % F removal with temperature (60 min contact time, 1.5 g dosage, 100 mL, 250 rpm, pH = 6.111).

Figure 5

(a) Variation of % F removal with contact time (1, 2 and 3 g, 10 mg/L F, 25 °C room temperature, 100 mL, 250 rpm, pH = 6.423). (b) Variation of % F removal with adsorbent dosage, (60 min contact time, 10 mg/L F, 25 °C room temperature, 250 rpm, pH = 5.937). (c) Variation of % F removal with initial concentration (60 min contact time, 2 g dosage, 25 °C room temperature, 100 mL, 250 rpm, pH = 6.111). (d) Variation of % F removal with temperature (60 min contact time, 1.5 g dosage, 100 mL, 250 rpm, pH = 6.111).

Effect of adsorbent dosage

To determine the minimum amount of adsorbent required to bring the fluoride concentration to the prescribed levels, the effect of adsorbent dose on fluoride removal efficiency was evaluated. Generally, the percentage removal of fluoride was observed to increase with dosage. As dosages increases, more adsorption sites suitable for the uptake of fluoride become available. From Figure 5(b), it can be seen that after 1.5 g, the fluoride removal remained fairly constant, indicating that the reaction had reached equilibrium. At low adsorbent dosage the fluoride adsorption rate is rapid since the active sites are easily available and at high adsorbent dosage, the adsorbate species find it increasingly difficult to access the adsorption sites and equilibrium is established (Goswami & Purkait 2012; Balouch et al. 2013). Vinati et al. (2015) also revealed that decreased adsorption capacity with increasing adsorbent dose may be attributed to the reduced driving force for adsorption as high adsorbent dose causes lower equilibrium fluoride concentration.

Effect of initial fluoride concentration

Generally, the adsorption capacity of the adsorbent increases by increasing the initial concentration of the adsorbate molecules. As the adsorbate concentration increases, the number of collisions between the adsorbate molecules and the adsorbent functional groups also increases, leading to an increase in the adsorption capacity (Guo & Reardon 2012; Ismail & AbdelKareem 2015). The effect of initial concentration of fluoride on the adsorption and uptake of fluoride from solution by MMCS was evaluated by varying the fluoride concentration solutions from 1 to 15 mg/L. The percentage fluoride removal of the adsorbent decreased with increasing initial fluoride concentration whereas the adsorption capacity of the adsorbent increased with increasing initial fluoride concentration. Some studies have also reported similar results (Daifullah et al. 2007; Cai et al. 2016). This may be due to the number of active sites on the adsorbent that had not reached saturation at lower fluoride concentrations. The maximum adsorption capacity was observed at 9 mg/L.

Effect of temperature

Temperature usually has a notable effect on reaction rate. If temperature increases, the rate of chemical reaction also increases. As the temperature was increased, fluoride removal was also observed to increase gradually (Figure 5(d)). However the increase in fluoride percentage removal is minor on all three tested temperatures. From the obtained results, it was noted that increase in temperature had a negligible impact on fluoride removal. Izuagie et al. (2016) also reported an insignificant percent fluoride removal and adsorption capacity for different concentrations of fluoride at the different evaluated temperatures. The adsorbent would therefore be applicable for household defluoridation of groundwater at the ambient temperature.

Effect of pH

The pH of a medium is one of the most important parameters that can influence fluoride removal. It also helps in understanding the fluoride uptake mechanism of the adsorbent. As shown in Figure 6, the adsorption decreases with increase in pH. At low pH the surface of the adsorbent will be positively charged, enhancing F adsorption. An interaction between the metal hydroxides (Si-OH, Al-OH and Fe-OH) at the surface of the clay adsorbent and fluoride ions is shown by Equations (3) and (4). The specific adsorption of fluoride on the clay surface (S) follows the outline:
Figure 6

Percentage removal of fluoride by MMCS as a function of pH (60 min contact time, 1.5 g, dosage, 9 mg/L F, 25 °C room temperature, 100 mL, 250 rpm).

Figure 6

Percentage removal of fluoride by MMCS as a function of pH (60 min contact time, 1.5 g, dosage, 9 mg/L F, 25 °C room temperature, 100 mL, 250 rpm).

At acidic pH 
formula
3
 
formula
4
whereas at higher pH (basic condition), there is dissolution of metal hydroxides from the adsorbents and this leads to increase in OH in the water.
At alkaline pH 
formula
5
The progressive decrease of fluoride uptake with increase in pH is mainly due to the electrostatic repulsion of fluoride ions by the negatively charged surface of the adsorbent and competition of adsorption sites with OH, since both OH and F are isoelectronic and have the same charge and similar ionic radii. Thus the adsorption of fluoride ion followed ligand exchange mechanism.

Regeneration tests

Regeneration experiments were carried out to evaluate the reusability of MMCS. The regeneration for MMCS is presented in Figure 7. It can be seen that the regeneration rate of the clay descended with progressing desorption–adsorption cycles. After four regeneration cycles, MMCS still showed a significant fluoride removal. It showed that the clay could be well regenerated by 0.01 mol/L NaOH.
Figure 7

Regeneration test of MMCS (60 min contact time, 2 g dosage, 9 mg/L F, 25 °C room temperature, 100 mL, 250 rpm).

Figure 7

Regeneration test of MMCS (60 min contact time, 2 g dosage, 9 mg/L F, 25 °C room temperature, 100 mL, 250 rpm).

Chemical stability of MMCS

From Table 4, it can be observed that higher leaching of metals occurred at pH 12. However with Al, Mn and Fe leaching occurred at all the tested pHs. When compared to the South African water quality guidelines the metal Al exceeds the recommended limits at those different pHs, but at pH 6 the leaching of metals is quite low. However, according to the Department of Water Affairs and Forestry (DWAF), at those concentrations, the metals have no acute health effects (DWAF 1996). Contrary to other metals, Na shows a quite different trend. Na concentrations should range between 0 and 100 mg/L (DWAF 1996) and in Table 4, the Na concentration at pH 12 is 272.7 mg/L, which is above the recommended limit. The high Na concentration came from the NaOH which was used to adjust the pH to 12. Since at this pH, defluoridation was poor, the pH will not be used for defluoridation. Since the defluoridation experiments avoid secondary pollution for most of the metals across a wide and useful pH range, MMCS can be considered a promising material for fluoride removal from drinking water.

Table 4

Leaching of metals into solution after defluoridation at pH 2, 6 and 12

Metals (μg/L)2612
Li 3.54 4.33 6.31 
Be 2.93 0.27 1.60 
4.46 2.61 <0.000 
Al 5,681.6 3,851.3 17,337.4 
15.83 4.70 66.38 
Cr 14.05 23.94 171.75 
Mn 1,405.1 44.11 396.36 
Fe 112.866 2,486.13 16,615.10 
Co 43.15 1.70 18.45 
Ni 220.08 19.59 166.86 
Cu 24.88 2.86 54.54 
Zn 25.55 5.41 40.26 
As 0.11 0.08 0.39 
Se 0.73 0.45 2.76 
Sr 169.07 6.99 35.88 
Mo 0.07 0.25 0.14 
Cd 0.23 0.01 0.05 
Sb 0.01 0.01 0.02 
Ba 424.07 7.26 124.48 
Hg <0.01 0.01 0.01 
Pb 1.62 0.42 5.94 
Ca 22.95 1.431 6.231 
2.465 0.9728 2.236 
Mg 20.33 1.314 9.39 
Na 12.99 9.353 272.7 
0.0394 0.0166 0.068 
Si 5.49 8.216 40.62 
Metals (μg/L)2612
Li 3.54 4.33 6.31 
Be 2.93 0.27 1.60 
4.46 2.61 <0.000 
Al 5,681.6 3,851.3 17,337.4 
15.83 4.70 66.38 
Cr 14.05 23.94 171.75 
Mn 1,405.1 44.11 396.36 
Fe 112.866 2,486.13 16,615.10 
Co 43.15 1.70 18.45 
Ni 220.08 19.59 166.86 
Cu 24.88 2.86 54.54 
Zn 25.55 5.41 40.26 
As 0.11 0.08 0.39 
Se 0.73 0.45 2.76 
Sr 169.07 6.99 35.88 
Mo 0.07 0.25 0.14 
Cd 0.23 0.01 0.05 
Sb 0.01 0.01 0.02 
Ba 424.07 7.26 124.48 
Hg <0.01 0.01 0.01 
Pb 1.62 0.42 5.94 
Ca 22.95 1.431 6.231 
2.465 0.9728 2.236 
Mg 20.33 1.314 9.39 
Na 12.99 9.353 272.7 
0.0394 0.0166 0.068 
Si 5.49 8.216 40.62 

Treatment of field water from Siloam borehole at optimized pH conditions

From Table 5, the fluoride removal efficiency on borehole water at field pH conditions was found to be slightly lower than that at optimized pH conditions. A possible explanation for this could be that the borehole water had a circumneutral pH and according to Figure 6 which shows the effect of pH on fluoride removal, at that pH, percentage fluoride removal is very low. However, at both field and optimized pH conditions fluoride levels were above permissible limits of 1.5 mg/L. This may be due to the fact that that the fluoride removal capacity of MMCS is fairly low. The other parameters are however within the permitted ranges of the DWS water quality guidelines.

Table 5

Adsorption of fluoride onto MMCS under field pH and optimized pH conditions

ParameterSiloam borehole waterField pH (pH = 7.81)Optimized pH (pH = 2.12)DWS guidelines
pH 7.037 7.091 6.424 6–9 
EC (μS/cm) 25.74 25.22 25.20 0–150 
F (mg/L) 5.53 3.233 2.95 1–1.5 
Cl(mg/L) 31.6 38.1 67.1 0–250 
(mg/L) 11.9 10.5 9.6 0–250 
ParameterSiloam borehole waterField pH (pH = 7.81)Optimized pH (pH = 2.12)DWS guidelines
pH 7.037 7.091 6.424 6–9 
EC (μS/cm) 25.74 25.22 25.20 0–150 
F (mg/L) 5.53 3.233 2.95 1–1.5 
Cl(mg/L) 31.6 38.1 67.1 0–250 
(mg/L) 11.9 10.5 9.6 0–250 

Adsorption isotherms

The analysis of equilibrium data helps to develop mathematical models that could be used for the quantitative description of the results. The equation parameters and the underlying assumptions of these equilibrium models are capable of predicting ion adsorption and vital information on the mechanism of sorption. The Langmuir and Freundlich isotherms were tested in this study and their parameters are given in Table 6.

Table 6

Adsorption isotherm parameters

  Langmuir Isotherm
 
Temp (K)Qm (mg/g)b (L/mg)RLR2
298 0.18 1.74 0.06 0.5203 
308 0.11 3.25 0.03 0.3107 
328 0.11 1.9 0.06 0.2893 
  Freundlich Isotherm
KF1/nnR2
298 0.06 3.33 0.2997 0.9529 
308 0.07 5.23 0.191 0.9687 
328 0.09 6.03 0.1658 0.9558 
  Langmuir Isotherm
 
Temp (K)Qm (mg/g)b (L/mg)RLR2
298 0.18 1.74 0.06 0.5203 
308 0.11 3.25 0.03 0.3107 
328 0.11 1.9 0.06 0.2893 
  Freundlich Isotherm
KF1/nnR2
298 0.06 3.33 0.2997 0.9529 
308 0.07 5.23 0.191 0.9687 
328 0.09 6.03 0.1658 0.9558 

The Langmuir isotherm Equation (6) assumes that adsorption cannot proceed beyond the monolayer and the ability of a molecule to be adsorbed at a given site is independent of the occupation of neighbouring sites.

Expressed mathematically 
formula
6
where Ce is the equilibrium concentration (mg/L), Qe is the amount adsorbed at equilibrium (mg/g), b represents the Langmuir isotherm constant and Qm is the maximum adsorption capacity for a complete monolayer coverage.
The values of the linear regression equation (R2), Qm and b are shown in Table 6. From the R2 value obtained (Figure 8), it is shown that the Langmuir isotherm provided a poor fit to the experimental data. The constant b is related to the affinity between the adsorbent and adsorbate. A low value of b indicates favourable adsorption. The RL value indicates the shape of the isotherm and is given in Equation (7). 
formula
7
The values of RL obtained at different temperatures are shown in Table 6. The values lie between 0 and 1 which shows that the adsorption of F on Mukondeni clay soils is favourable. The calculated adsorption capacities (Qm) shown in Table 6 are very high compared to the experimental adsorption capacity (0.08 mg/g). This is also another indication that the Langmuir isotherm does not describe the adsorption process that took place.
Freundlich isotherm Equation (8) is mainly used to describe the adsorption characteristics for heterogeneous surfaces (Krishnaiah & Vijaya 2008).
Figure 8

Langmuir (a) and Freundlich (b) adsorption isotherm for fluoride adsorption onto MMCS (2 g adsorbent, 60 min of equilibration time, concentration was varied from 1 to 15 mg/ L, pH = 6,111).

Figure 8

Langmuir (a) and Freundlich (b) adsorption isotherm for fluoride adsorption onto MMCS (2 g adsorbent, 60 min of equilibration time, concentration was varied from 1 to 15 mg/ L, pH = 6,111).

Expressed mathematically 
formula
8
KF and 1/n are the Freundlich constants, describing the adsorption capacity and intensity respectively. Table 6 shows the Freundlich isotherm parameters. The high R2 values (Figure 8) obtained clearly suggest that the applicability of the Freundlich adsorption isotherm is feasible and that the adsorption process occurred on a heterogeneous surface. Furthermore the calculated adsorption capacities (0.06, 0.07 and 0.09 mg/g) are very close and comparable to the experimental adsorption capacity (0.08 mg/g), substantiating that the Freundlich adsorption isotherm best describes the adsorption of fluoride onto Mukondeni clay soils.

Adsorption kinetics

The kinetics of adsorption controls the efficiency of the process and the equilibrium time. To identify the potential rate controlling steps involved in the process of adsorption, two kinetic models were studied and utilized to fit the experimental data from the adsorption process.

The pseudo-first-order equation

The pseudo-first-order equation (Lagergren's equation) describes adsorption in solid–liquid systems based on the sorption capacity of solids. The linear form can be expressed as: 
formula
9
where qe (mg/g) is the adsorption capacity at equilibrium, qt (mg/g) is the adsorption capacity at time t, and K (1/min) is the rate constant of pseudo-first-order. The value of K can be obtained from the slope by plotting log (qeqt) vs t.

Pseudo-second-order kinetics

The pseudo-second-order rate expression, which is applied for analyzing chemisorption kinetics from liquid solutions pseudo-second based on solid phase sorption is linearly expressed as: 
formula
10
where is the rate constant for pseudo-second-order adsorption (g/mg/h) or h (mg/g/h) is the initial adsorption rate.

The pseudo-first-order kinetics best fit line yielded relatively low R2 values. This suggests that the application of the pseudo-first-order is inappropriate hence indicating that the adsorption of fluoride onto Mukondeni clay soils did not follow pseudo-first-order kinetics. On the other hand, the correlation coefficient (R2) for the pseudo-second-order kinetic model fit is ≈1 as shown in Table 7. Given the good agreement between model fit and experimentally observed equilibrium adsorption capacity in addition to the large correlation coefficients, it can be suggested that fluoride adsorption followed pseudo-second-order kinetics and F ions were adsorbed onto the Mukondeni clay soils surface via a chemical interaction.

Table 7

Pseudo-second-order kinetic parameters for different adsorbent dosages

Dosage (g)qe exp (mg/g)qe cal (mg/g)K2(gmh)R2
0.15 0.078 2.02 0.9933 
0.12 0.083 2.52 0.9959 
0.08 0.057 2.91 0.9915 
Dosage (g)qe exp (mg/g)qe cal (mg/g)K2(gmh)R2
0.15 0.078 2.02 0.9933 
0.12 0.083 2.52 0.9959 
0.08 0.057 2.91 0.9915 

Performance comparison of Mukondeni clay soils with other clay adsorbents

A substantial amount of literature is available on fluoride sorption by a wide variety of low cost adsorbents, as shown in Table 8. A summary of adsorption capacities of various adsorbents for fluoride removal from water under various experimental conditions has been presented in Table 8. A perusal of Table 8 reveals that mineral-based sorbents have been found to be promising for fluoride removal. As can be seen from the Table 8, MMCS have the lowest adsorption capacity when compared to other materials that have been previously used for adsorption but the highest adsorption capacities are observed with modified clays.

Table 8

Comparison of different adsorption capacities of different clay adsorbents for fluoride

AdsorbentAdsorption capacityExperimental conditionsReferences
Mukondeni clay soils 0.08 mg/g pH 2; 9 mg/L Present study 
Pyrophyllite 0.737 mg/g pH 4–9; 10 mg/L Kim et al. (2013)  
Mg2+ bentonite 2.3 mg/g pH 3–10; 5 mg/L Thakre et al. (2011)  
Montmorillonite 1.324 mg/g pH 4; 1–6 mg/L Ramdani et al. (2010)  
Pyrophyllite 2.2 mg/g pH 4–9; 10 mg/L Goswami & Purkait (2011)  
AdsorbentAdsorption capacityExperimental conditionsReferences
Mukondeni clay soils 0.08 mg/g pH 2; 9 mg/L Present study 
Pyrophyllite 0.737 mg/g pH 4–9; 10 mg/L Kim et al. (2013)  
Mg2+ bentonite 2.3 mg/g pH 3–10; 5 mg/L Thakre et al. (2011)  
Montmorillonite 1.324 mg/g pH 4; 1–6 mg/L Ramdani et al. (2010)  
Pyrophyllite 2.2 mg/g pH 4–9; 10 mg/L Goswami & Purkait (2011)  

CONCLUSIONS AND RECOMMENDATIONS

Mukondeni clay soils were found to be a good and cheap adsorbent for the removal of fluoride ions from aqueous solution but could be further modified to increase their adsorption capacity. To the best of our knowledge, MMCS were never tested for water defluoridation technologies and this study has proved that they have water treatment capabilities and hence this is a novel finding in clay based fluoride adsorbents. This information will also be useful to the rural communities at large because MMCS can be modified to be used for defluoridation purposes. Solution pH played a major role in fluoride removal, it influenced the positive charge on the adsorbent and hence improved adsorption capacity with the pH of 2 being recorded for highest fluoride removal but minimal metal leaching was exhibited across a wide range of pH 4–10. The Freundlich isotherm gave a much better fit than the Langmuir isotherm indicating that the adsorption occurred on a heterogeneous surface. Fluoride adsorption followed pseudo-second-order kinetics and F ions were adsorbed onto the MMCS surface through a chemical interaction. The adsorption mechanism of adsorbents for fluoride could involve anion exchange and electrostatic interaction based on PZC and FTIR results. MMCS have shown fairly good fluoride uptake capacity in the laboratory. It is recommended that a clay composite be made between MMCS and metal oxide modified clay to enhance the adsorption capacity for defluoridation purposes. It is further recommended that the clay soils disposed of after defluoridation experiments be used to make decorative clay pottery as the clay soils were initially used for these purposes, and if the clay soils are in large quantities they can be used for brick making.

REFERENCES

REFERENCES
Balouch
A.
Kolachi
M.
Talpur
F. N.
Khan
H.
Bhanger
M. I.
2013
Sorption kinetics, isotherm and thermodynamic modeling of defluoridation of groundwater using natural adsorbents
.
Am. J. Anal. Chem.
4
,
221
228
.
Bhatnagar
A.
Kumar
E.
Sillanpaa
M.
2011
Fluoride removal from water by adsorption-a review
.
Chem. Eng. J.
10
,
5
28
.
Cai
H.
Xu
L.
Chen
G.
Peng
C.
Ke
F.
Liu
Z.
Li
D.
Zhang
Z.
Wana
X.
2016
Removal of fluoride from drinking water using modified ultrafine tea powder processed using a ball-mill
.
Appl. Surf. Sci.
375
,
74
84
.
Chapman
H. D.
1965
Cation exchange capacity
. In:
Methods of Soil Analysis – Chemical and Microbiological Properties
(
Black
C. A.
, ed.).
Agronomy, 9, 891–901. California Digital Library, California
.
Denga
M. E.
2013
Evaluation of the Fluoride Adsorption Capacity of Locally Available Clay in Vhembe District
.
Unpublished
.
University of Venda
,
Thohoyandou, South Africa
.
DWAF
1996
South African water quality guidelines, vols. 3 (Industrial use) and 4 (Commercial irrigation)
, 2nd edn.
Department of Water Affairs and Forestry
,
Pretoria
.
Gitari
W. M.
Ngulube
T.
Masindi
V.
Gumbo
J. R.
2013
Defluoridation of groundwater using Fe3+-modified bentonite clay: optimization of adsorption conditions
.
Desalin. Water Treat.
53
,
1578
1590
.
Goswami
A.
Purkait
M. K.
2012
The defluoridation of water by acidic alumina
.
Chem. Eng. Res. Des.
90
,
2316
2324
.
Goswami
A.
Purkait
M. K.
2013
Defluoridation of water by schwertmannite
.
World Academy of Science, Engineering and Technology
73
,
1156
1161
.
Izuagie
A. A.
Gitari
W. M.
Gumbo
J. R.
2016
Synthesis and performance evaluation of Al/Fe oxide coated diatomaceous earth in groundwater defluoridation: towards fluorosis mitigation
.
J. Env. Sci. Health A
1
15
,
810–824
.
Kim
J. H.
Lee
G.
Park
J. A.
Kang
J. K.
Choi
N. C.
Kim
S. B.
2013
Use of pyrophyllite clay for fluoride removal from aqueous solution
.
Desalin. Water Treat.
51
,
3408
3416
.
Krishnaiah
A.
Vijaya
Y.
2008
Sorptive response profile of chitosan coated silica in the defluoridation of aqueous solution
.
E-Journal of Chemistry
6
(
3
),
713
724
.
Langmuir
D.
1997
Aqueous Environmental Geochemistry
.
Prentice-Hall, Inc.
,
New Jersey
.
Thakre
D.
Rayalua
S.
Kawadea
R.
Meshramb
S.
Subrtc
J.
Labhsetwara
N.
2011
Magnesium incorporated bentonite clay for defluoridation of drinking water
.
J. Hazard. Mater.
180
,
122
130
.
Vázquez-Guerreroa
A.
Alfaro-Cuevas-Villanueva
R.
Rutiaga-Quĩnones
J. G.
Cortés-Martínezd
R.
2016
Fluoride removal by aluminum-modified pine sawdust: effect of competitive ions
.
Ecol. Eng.
94
,
365
379
.
WHO
2011
Chemical Fact Sheets. Guidelines for Drinking-water Quality
, 4th edn.
Gutenberg
,
Malta
, pp.
370
373
.
WRC
2001
Distribution of Fluoride-Rich Groundwater in the Eastern and Mogwase Regions of the Northern and North-West Province
.
WRC Report No. 526/1/01 1.1 - 9.85
.
Pretoria
,
South Africa
.