Introduction

Fluoride is a normal constituent of natural water because of its high reactivity. Normally, fluorine exists in the form of fluoride in natural waters (Leung and Hrudey 1985). Its concentration, though, varies significantly depending on the water source. Although both geological and manmade sources contribute to the occurrence of fluoride in water, the major contribution comes from geological sources like presence of fluoride-rich minerals (Kumar et al. 2016). The most common fluoride-containing minerals are fluorspar, cryolite, muscovite, biotite, fluorite and fluorapatite (Avtar et al. 2013). Along with the geological and mineralogical signature of the aquifer matrix, other physico-chemical characteristics of the aquifer such as porosity and alkalinity of the soil and rocks, temperature, chemical reaction of co-existing ions and the depth of wells also play a contributing role for release of fluoride in groundwater. In India, fluoride is the major inorganic pollutant of natural origin found in groundwater and, due to large number of variables, the fluoride concentrations in groundwater range from well under 1.0 mg/L to more than 35.0 mg/L (IPCS 1984; Kumar et al. 2016). From human health perspectives, fluoride helps in the normal mineralization of bones and formation of dental enamel (Cao et al. 2000; Rajkumar et al. 2015). The total daily intake of fluoride from food is about 0.2–0.5 mg which is about only 10–15 % of the required dose and hence we have to be dependent on groundwater to fulfil this deficit (Boyle and Chagnon 1995). The desirable safe limit of fluoride in drinking water is 1.0 mg/L (WHO 1984). In case the daily intake of fluoride is low (i.e. <0.5 mg/L), various health issues may occur, viz. dental caries, lack of formation of dental enamel and deficiency of mineralization of bones, especially affecting children (Ingle et al. 2014). On the other hand, excess daily intake of fluoride (i.e. >1.5 mg/L) also causes several health-related problems, viz. fluorosis affecting all age group people (Mondal et al. 2009). When fluoride is consumed in the range of 1.5–2.0 mg/L, dental fluorosis or dental mottling may occur, characterized by brown or black opaque patches on the enamel/tooth surface (Kharb and Susheela 1994). Intake of fluoride exceeding 3.0 mg/L for a longer period of time results in skeletal fluorosis characterized by deformation of bones (Goldman et al. 1991). Other than the above-mentioned diseases, excessive intake than the recommended limit of fluoride may lead to increased thirst, skin rashes, muscle fibre degeneration, blood cell deformation, gastrointestinal problems, urinary tract malfunctioning, and overall reduced immunity(Meenakshi and Maheshwari 2006; Singh et al. 2011). At the global scale, higher concentrations of fluoride (i.e. >1.5 mg/L) in groundwater and related health effects is well reported in more than 30 countries, namely China, Syria, Jordan, Ethiopia, Sudan, Tanzania, Kenya, and Uganda (Ando et al. 2001; Razbe et al. 2006). In the case of India, groundwater contamination with fluoride is well reported at numerous places in the states of Andhra Pradesh, Gujarat, Karnataka, Madhya Pradesh, Rajasthan, Chhattisgarh, Haryana, Orissa, Punjab, Haryana, Uttar Pradesh West Bengal, Bihar, Delhi, Jharkhand, Maharashtra, and Assam (Keshari and Dhiman 2001; Jacks et al. 2005; CGWB 2010). In Gujarat, the Government of India has highlighted Mehsana District in particular as a water quality concern area with specific reference to fluoride enrichment. Many studies reveal that infants, children and adults in Mehsana District are exposed to high doses of fluoride from groundwater (Chinoy et al. 1992; Dhiman and Keshari 2003).

The fluoride contamination and its removal approach attract a lot of concern from the scientific community at the global scale due to finite potable groundwater resource. The main traditional techniques used by the scientific community for defluoridation are ion exchange, immobilization, electrodeposition, membrane separation, precipitation and adsorption. Amidst these techniques, adsorption is a widely accepted method for defluoridation because of its ease of operation as well as economic feasibility (Ingle et al. 2014). Chidambaram et al. (2013) gave a detail literature review for the different adsorption materials (both natural and synthetic) used for defluoridation from groundwater. Along with that, some of the adsorbents used recently are microwave-assisted activated carbon (Dutta et al. 2012), physico-chemically treated sand (Togarepi et al. 2012), pumice (Malakootian et al. 2011) and raw bauxite (Sajidu et al. 2012). The most commonly reported adsorbents are activated alumina and activated carbon. However, treatment and disinfection of water for drinking purpose using available mitigation approaches make it too expensive and complex for application in poor communities.

As mentioned above, although there are few scientific publications targeting defluoridation by natural adsorbent, targeting adsorbent on its availability and affordability is still very sparse and more attention needs to be paid. Therefore, in this study an effort is made to observe the adsorption of fluoride from groundwater of Mehsana District, Gujarat, India, on finely chopped rice husk, powdered rice husk and sawdust. These adsorbents are also chosen on the basis of availability (very commonly available) as well as affordability (very cheap) with local residents there.

Study area

Mehsana District of Gujarat is located between 23°15′ and 23°53′N latitudes and 72°07′ and 72°26′E longitude and shares a common border with Patan District (Fig. 1). It has a total geographical area of 4393 km2 which is divided into nine subsections called talukas. The climate of the area is warm, sub-humid, sub-tropical and monsoonal. The month of May is generally the hottest and January is considered the coldest. The mean, maximum and minimum temperature is 27, 45 and 15 °C, respectively. The mean annual rainfall is 600 mm and about 93.5 % of the total rainfall is received during the month of June to September by the south-west monsoon and maximum rainfall is received during the month of July and August. The area is intersected by numerous streams, namely Sabarmati, Rupen, Saraswati, Khari, and Pushpawati. The geomorphology of the area is almost monotonously flat and featureless and is represented by alluvial plain.

Fig. 1
figure 1

Study area map showing sampling location

Full size image

Sampling and analytical techniques

Based on the consumer’s health issues recorded as shown in (Fig. 2), a total of 34 samples were collected from the study area. Most of the groundwater samples were collected from tube wells. The sample coordinates were recorded using the global positioning system (GPS III, Garmin). Groundwater samples were collected using clean polyethylene bottles, and water samples were collected after pumping the water for 5–10 min. Following the collection, samples were brought to the laboratory in an ice chest and stored at below 4 °C for fluoride analysis. Fluoride was analysed by the SPANDS method using the Jenway model 6505 spectrophotometer (APHA 1995), and high-purity reagents (Merck) and Milli-Q water (Model Milli-Q, Biocel) were used for all the analysis. Analytical reagent-grade sodium fluoride (NaF), SPADNS reagent (4, 5-dihydroxy-3-(P-sulfophenylazo)-2, 7-naphthalene-disulfonic acid trisodium salt), zirconium oxychloride, concentrated HCl, and distilled water were used in the analysis of fluoride. A stock solution of fluoride was prepared by dissolving sodium fluoride in distilled water and working fluoride solution of different concentrations were prepared from stock fluoride solution by appropriate dilution.

Fig. 2
figure 2

Different health ailments in the study area related to consumption of fluoride-contaminated water. a Skeletal and b, c dental fluorosis

Full size image

All the experiments of batch adsorption are piloted to examine the effect of various parameters like dose of adsorbent, pH of working media, initial concentration of fluoride in the working sample, and contact time for which the adsorbent was kept in the working media. All experiments are conducted in a closed chamber at constant background room temperature of 25 ± 2 °C measured with laboratory-installed thermometer. During the whole reaction time, the conical flask was kept on a shaker with constant speed at 130 rpm (rotations per minute). On completion of reaction time, the adsorbate is separated from the solution using Whatman filter paper No. 42 and the filtrate is analysed for residual fluoride concentration using the SPADNS method. A known weight of adsorbent material is added into 50 mL of prepared samples of fluoride taken in a conical flask.

The effect of adsorbent dose on defluoridation is calculated by adding 0.5, 1, 2, 3, 4, 5 and 10 g of the adsorbent in a working solution of known concentration (5 mg/L) taken in seven different flasks. After 10 h of reaction time, the fluoride concentration is measured. The fluoride adsorption from solution is strongly influenced by the pH of the working solution and the effect of the pH of the working solution on fluoride adsorption is studied by adjusting the pH of the working solution from pH 4.0 to 8.0. The initial concentration of the targeted pollutant also has remarkable effect on its removal by adsorption. Thus, samples of different fluoride concentrations (2, 5, 10, 15, 20, 25 and 50 mg/L) are prepared and 2 g of adsorbent added in all the working samples. The effect of contact time between the adsorbent and adsorbate for effective fluoride removal was studied by preparing the samples of known concentration and adding 2 g of adsorbent and conducts the experiment for same time period. To study the effect of time on fluoride removal, concentration of fluoride was measured at an interval of 1 h initially up to 8 h, followed by at an interval of 2 h up to 16 h.

Kinetic studies of sorbent were carried out in a temperature-controlled mechanical shaker. The effect of different initial fluoride concentrations, viz., 2, 4, 6, 8 and 10 mg/L at 25 °C temperature on sorption rate, was studied by keeping the mass of sorbent as 2 g and volume of solution as 50 mL in neutral pH. The fluoride concentration retained in the adsorbent phase, q e (mg/g), was calculated according to Eq. (1):

$$q_{\text{e}} = \frac{{\left( {C_{\text{o}} - C_{\text{e}} } \right)}}{W},$$
(1)

where q e is the amount of fluoride adsorbed (mg/g); C o and C e are the initial and residual fluoride concentration in solution at equilibrium (mg/L) respectively; and W is the weight (g) of the adsorbent.

Adsorption equilibrium data were examined with the most widely used being the theoretical Langmuir isotherms model (Langmuir 1916). It is often applied in solid/liquid system to describe the saturated monolayer adsorption well represented by Eq. (2):

$$q_{\text{e}} = \frac{{q_{\text{m}} K_{\text{a}} C_{\text{e}} }}{{1 + K_{\text{a}} C_{\text{e}} }},$$
(2)

where C e is the equilibrium concentration (mg/L); q e is the amount of ion adsorbed (mg/g); q m is q e for a complete monolayer (mg/g); K a is adsorption equilibrium constant (L/mg). To evaluate the adsorption capacity for a particular range of adsorbate concentration, the above Eq. (2) can be used in linear form as mentioned by Eq. (3):

$$\frac{{C_{\text{e}} }}{{q_{\text{e}} }} = \frac{1}{{q_{\text{m}} }}C_{\text{e}} + \frac{1}{{K_{\text{a}} q_{\text{m}} }}.$$
(3)

The constants q m and K a can be determined from a linearized form of Eq. (2) by the slope of the linear plot of C e/q e versus C e.

Results and discussion

A complete statistics of fluoride distribution pattern in the study area is shown in Table 1. From geochemical analysis of groundwater samples, it is found that fluoride concentration ranges from 0.32 to 4.79 mg/L. More than 70 % of groundwater samples are found with fluoride concentration exceeding from 1.0 mg/L. pH of the water samples ranges from 7.2 to 7.7 with an average value of 7.4, indicating that groundwater is slightly alkaline in nature.

Table 1 Statistical analysis of fluoride contamination in groundwater samples for Mehsana District
Full size table

Result for fluoride adsorption

Here in this section, the effect of different parameters on the rate of defluoridation is assessed and described as follows:

The effect of adsorbent dose

To estimate the optimum dose for the effective fluoride adsorption, it is an essential task to find the quantity of adsorbent that is adequate. The experiment to evaluate the effect of dose on adsorption is carried out with an addition of 0.5, 1, 2, 3, 4, 5 and 10 g in the laboratory-prepared sample with an initial concentration of fluoride of 10 mg/L. The result for percentage fluoride adsorption with dose of adsorbents is shown in Fig. 3. It is found that up to a certain extent, with increase in the amount of dose of adsorbent, the percentage of fluoride adsorption also increases. However after that certain amount of dose, no further significant increase in fluoride adsorption is observed. Finally, it is concluded that the lowest quantity of adsorbent required for maximum adsorption was 2 g of adsorbent per 50 mL of adsorbate.

Fig. 3
figure 3

Scatter plot showing the relation between percentage adsorption of fluoride and dose of adsorbent

Full size image

Effect of pH

The effect of pH on the percentage adsorption of fluoride from the adsorbent surface is estimated by varying the pH of the working solution from 4 to 8 at an adsorbent dose of 2 g in 50 mL of solution with an initial concentration of 10 mg/L. The result for percentage adsorption of fluoride versus pH of the working solution is shown in Fig. 4. It is found that a higher percentage of fluoride adsorption took place when the pH is in the acidic range with a maximum absorption at pH = 4 (Fig. 4).

Fig. 4
figure 4

Scatter plot showing the relation between the percentage adsorption of fluoride and pH of the working solution

Full size image

Effect of initial concentration

A given mass of adsorbent can adsorb only a fixed amount of adsorbate, so the initial concentration of the adsorbate solution is very important. The effect of the initial concentration of fluoride in water on the removal of fluoride was studied by varying the initial concentration of fluoride from 5 to 50 mg/L, keeping the optimized value of other parameters, viz., dose of adsorbent 2 g per 50 mL of solution and pH at 4. It is observed that the percentage fluoride removal increases with increase in the initial concentration of the solution, but the efficiency becomes stagnant or even decreases after a certain level. Also, it is clear that the effective percentage removal of fluoride for finely chopped rice husk, powdered rice husk and sawdust takes place at an initial concentration of 10, 10 and 15 mg/L, respectively (Fig. 5).

Fig. 5
figure 5

Scatter plot showing the effect of the initial concentration of fluoride in the sample on percentage adsorption by different adsorbents

Full size image

Effect of contact time

The effect of contact time for the removal of fluoride using finely chopped rice husk, powdered rice husk and sawdust was investigated by analysing the samples collected during treatment at different time intervals. The optimized values of other parameters such as adsorbent dose and pH, as mentioned above, were used. The initial concentration of the synthetic sample was taken as 10 mg/L and the experiment was done in a temperature-controlled shaker at 130 rpm and 25 ± 2 °C. The results for contact time versus percentage adsorption of fluoride are shown in Fig. 6. Here, it is evident that the rate of fluoride removal increases with time, but after some time reaches the optimum level called saturation condition, beyond which no further adsorption takes place. The experimental value of optimum contact time in case of finely chopped rice husk and powdered rice husk was 7 h, whereas for sawdust the optimum value was 12 h.

Fig. 6
figure 6

Scatter plot showing the relation between the percentage adsorption of fluoride and contact time with three different adsorbents

Full size image

Comparative study of defluoridation efficiency using optimized parameters (determined at laboratory scale) versus natural pH of groundwater sample in the study area

A comparative study was conducted to assess the defluoridation efficiency of the above optimized parameter (including pH) versus optimized parameters (excluding pH) using three different natural adsorbents at the field scale. The objective of this experiment was to look for cost minimization using different natural adsorbents. There are three places where we can adjust the cost of defluoridation:

Adsorption dose

The optimum dose for significant defluoridation for three different adsorbents studied here was 2 g (Fig. 3). As plenty of all three adsorbents were locally available as well as it is really cheap (ranging from 0.0045 to 0.045 USD/kg); so it should not be of major concern.

pH

Optimum pH for defluoridation using all three adsorbents was pH = 4 (Fig. 4), but it was also found that a significant amount of fluoride reduction took place at pH = 7. To lower down the pH of groundwater samples (which are slightly alkaline in nature) to acidic range, a lot of money has to be pumped in the form of different chemicals by the local consumer, which do not appear sustainable or convincing for long periods of time. On the other hand, addition of chemicals to reduce pH might also alter the water quality, which also raises the question on authentication of this technique.

Contact time

Although optimum time for significant defluoridation using different adsorbents was different and varied from 7 to 12 h (Fig. 6), local people over there can perform this experiment for optimum contact time considering the fact that they do not have to pay much extra money for longer period of time. The result of comparative study of field application of all optimized parameters including pH versus optimized parameters excluding pH (here pH 7 is considered as natural) using three different natural adsorbents for the selected four water samples from the study area is shown in Fig. 7. Here, it is found that the maximum amount of fluoride adsorption was observed in the case of powdered rice husk followed by fine chopped rice husk and saw dust. Also, it indicates that the difference of fluoride adsorption at different pH values using other optimized parameters was not significant. Hence, it is wise to apply this methodology for all samples at field scale.

Fig. 7
figure 7

Comparative assessment for percentage adsorption of fluoride at pH 3 and 7 for selected water samples (keeping optimized value fixed for dose and contact time). a, b, c Results for powdered rice husk, sawdust and finely chopped rice husk, respectively

Full size image

Adsorption isotherms

The equilibrium data isotherm analysis for fluoride adsorption onto the powdered rice husk at 25 °C temperature and neutral pH is shown in Fig. 8. Here, it is found that the adsorbent has a high affinity for fluoride adsorption under the given conditions (with r 2 = 0.9843). The related parameters obtained by calculation from the values of slopes and intercepts of the linear plot are shown in Table 2. The essential features of the Langmuir isotherm shape can be expressed in terms of a dimensionless constant separation factor or equilibrium parameter (R L), indicating whether an adsorption system is favourable or unfavourable, as defined by Eq. 4) Tan et al. (2009):

Fig. 8
figure 8

Langmuir isotherms obtained by using the linear method for the adsorption of fluoride using powdered rice husk at a temperature of 25 °C

Full size image
Table 2 Isotherm parameters obtained using the linear method for the adsorption of fluoride onto using powdered rice husk at a temperature of 25 °C
Full size table
$$R_{\text{L}} = \frac{1}{{1 + K_{\text{a}} C_{0} }},$$
(4)

where R L is a dimensionless separation factor, C 0 the initial fluoride concentration (mg/L) and K a the Langmuir constant (L/mg). If the value of R L > 1, the isotherm will be unstable; when the value of R L = 1, the isotherm will be linear; when the value of 0 < R L < 1, the isotherm will be favourable and when the value of R L = 0 the isotherm will be irreversible.

The relationship between R L and C 0 to represent the essential features of the Langmuir isotherm for powdered rice husk at 25 °C is shown in Fig. 9. Here, the values of the calculated R L for the given range of fluoride concentration are found in the range of 0.07–0.31, which suggests the favourable adsorption of fluoride onto this adsorbent under the conditions used for the experiments.

Fig. 9
figure 9

Plot between separation factor (R L) values versus initial fluoride concentration derived by Langmuir constants using powdered rice husk at a temperature of 25 °C

Full size image

Effect of temperature on the defluoridation mechanism using rice husk and feasibility of this technique at the ground level

Gibbs energy (∆G°) for the adsorption mechanism can be represented by Eqs. 5 and 6:

$$\Delta G^\circ = \Delta H^\circ - T\Delta S^\circ ,$$
(5)
$$\Delta {\text{G}}^\circ = - RT { \ln }K_{\text{c}} ,$$
(6)

where T is the  temperature in °K, K c the equilibrium constant (q e/C e). Also, it can be represented as a ratio of equilibrium concentration of fluoride attached to rice husk (q e) compared to Van’t Hoff equation as equilibrium concentration of rice husk in solution (C e). Here, the negative value of ∆G° shows the spontaneous feasible nature of the adsorption process. The value of ∆H° and ∆S° can be deduced from the slope and intercept, respectively, of the plot between lnK c and 1/T. Looking into monthly variation of average minimum and maximum temperature in the study area (Fig. 10), the result for the plot between lnK c and 1/T is shown in Fig. 11. Here, the range of the temperature considered is 283–333 °K. It is found that with increasing temperature, the sorption capacity also increases. Similarly, ∆G° also increases with rising tempertaure.

Fig. 10
figure 10

Variation of monthly average high and low temperature in the study area

Full size image
Fig. 11
figure 11

Scatter plot between lnK c and 1/T showing the effect of temperature on fluoride adsorption

Full size image

Application of powdered rice husk for defluoridation at the field scale

From Fig. 7, it is clear that powdered rice husk shows maximum percentage removal of fluoride among all three natural adsorbents as well as it requires the least time for effective adsorption. Therefore, powdered rice husk was used for adsorption of fluoride from 34 groundwater samples collected from villages of Mehsana District maintaining optimized values of the dose of adsorbent and contact time. The result obtained from the above experiment is given in Fig. 12. Here, it is seen that the percentage of fluoride removal varies from 37.9 to 94.3 with an average value of 68.7.

Fig. 12
figure 12

Status of fluoride concentration in raw water sample (initial concentration) and after being treated with powdered rice husk (final concentration) at natural pH of water samples. Here, the optimized values for adsorbent dose (2 g/50 mL) and contact time (7 h) were considered

Full size image

Finally, a comparative study shows the different techniques being used for defluoridation and how this study has extended the scientific findings for field application of low-cost adsorbent (Table 3). It is very clear that rice husk is commonly available at very cheap rates, making it a potential candidate for use in future.

Table 3 Comparative summary of different techniques used for defluoridation
Full size table

Conclusion

Fluoride pollution in groundwater is a global concern as ingestion of water with fluoride concentration more than 1.5 mg/L may result in dental or skeletal fluorosis and, in the recent past, the defluoridation from groundwater has become a major thrust area of investigation for the scientific community. There are many defluoridation techniques available in the market such as adsorption (using both natural and artificial adsorbent), reverse osmosis, electrodialysis, ion exchange and membrane filtration. The economic feasibility of different methods has still not been evaluated. This work is an attempt to evaluate the defluoridation efficiency using three newly introduced materials (powdered rice husk, finely chopped rice husk and sawdust) as natural adsorbent based on their availability and economic feasibility. It was found that all the three materials have the property to adsorb the fluoride from groundwater, but the powdered rice husk was most efficient for keeping the optimized value of the dose of the adsorbent and contact time. The pH is also one of the main factors controlling the adsorption process and the present study compared the performance for different pH values, i.e. optimized pH and natural pH of the groundwater, as adjustment of pH will not be possible in villages. It is found that powdered rice husk can be efficiently applicable for defluoridation with the natural pH of groundwater at the field scale. The sorption of fluoride using this adsorbent followed Langmuir isotherms. Finally, for checking the feasibility of powdered rice husk in removing the fluoride from groundwater samples collected from Mehsana District, it was found that it lowers the concentration of fluoride to permissible limit in most of the samples. Thus, it can be concluded that rice husk can be used as a good adsorbent for removing the fluoride from groundwater in those area where its contamination is common.