Defluoridation using pinecone-based activated carbon: Adsorption isotherm, kinetics, regeneration, and co-ions effect investigation

,


INTRODUCTION
No doubt, fluoride is an essential element for the development of bones and teeth yet it is regarded as an element of "boon and bane".It is known to have healing characteristics, but exceeding its permissible limit of 1.5 mg L -1 (WHO guidelines) could ultimately result in serious health problems in the long run.The World Health Organization also recognizes fluoride as one of the drinking water contaminants, alongside arsenic and nitrate, that contribute to significant health issues on a large scale. 1 Some of the problems associated with fluoride are dental and skeletal fluorosis, brittle bones, Alzheimer's syndrome and thyroid disorder, neurological damage, low haemoglobin levels, gastrointestinal problems, and infertility. 2,3The global prevalence of fluorosis is estimated to affect approximately 70 million individuals. 4Fluoride enters water from various sources, including naturally occurring geological sources such as fluorine-containing rocks and minerals (e.g., granite, basalt, syenite, fluorite, biotites, and topaz) as well as industrial activities such as metallurgical, electrodeposition, phosphate production, electronics, brick and iron-works, and ceramics. 5,6The elevated concentrations of fluoride are prevalent in many water sources, particularly in regions such as China, Sri Lanka, Rift Valley countries in Africa, and India.In India, chronic fluoride levels of 12-18 mg L -1 have been reported with 17 states showing excessive fluoride levels. 2,7ince fluoride levels in drinking water is an issue of global health concern, it needs to be reduced below acceptable level to provide safe drinking water.][10][11] But, most of these techniques fail when it comes to the rural population because of their cost except the adsorption method.Adsorption is an ancient and widely used method because of its high efficacy, operational and budget-friendliness for the removal of fluoride and other pollutants as well from water . 2,12n this regard, adsorbent such as activated carbon from biomass origin has been widely used to investigate studies related to fluoride removal.Biomass-based carbon over the years has gained a significant place as an alternate adsorbent due to its high surface area and reactivity, low-cost, eco-friendly characteristics, and renewability. 3,13,14For instance, Gebrewold et al., 2019 synthesised activated carbon from biomass of rice husk and corn cob for the removal of fluoride from groundwater with an adsorption capacity of 7.9 mg g -1 and 5.8 mg g -1 ; 15 Fito et al., 2019 also used Catha edulis as a potential precursor for synthesising activated carbon for the removal of fluoride from water; 16 Jayashree et al., 2021 studied the removal of fluoride from water using activated carbon prepared from pods of Bauhinia variegata by both physical and chemical methods with BET surface area of 71.59 m 2 g -1 and 124.82 m 2 g -1 .They found that chemically prepared activated carbon was having a higher fluoride adsorption capacity (15.74 mg g -1 ) over physically prepared carbon by a difference of 5 mg g -1 . 17herefore, biomass materials such as pinecones that are often discarded as waste can be potentially utilised to obtain a value-added product like activated carbon to remove fluoride from water.Our group have been working on the preparation of activated carbon from locally available biomass-Manihot A c c e p t e d m a n u s c r i p t esculenta, Schima wallichii and Mucina sp.-for the removal of fluoride from water. 4,18,19hus, in this piece of work, we explored the pinecone of Pinus kesiya as the precursor for the synthesis of activated carbon because of its easy availability, abundance, and low cost for the removal of fluoride from water.Characterization of the adsorbent was analysed with various techniques to study the physicochemical properties, surface morphology, and functionalities to access the idealness of the adsorbent for fluoride removal.Further, motivated by the presence and effects of existing co-ions on the adsorption efficiency, the prepared adsorbent was applied for defluoridation studies in the presence of various co-ions using a batch method to understand the efficacy of the adsorbent in a real-time situation.Regeneration of the adsorbent was also explored to understand its recyclability.

Materials
Pinecones were collected locally, washed, and grounded using a mixer grinder to fine particles.All other chemicals used in this work were of analytical grade.

Synthesis of Pinecone activated carbon
For the synthesis of activated carbon, a single-step method published in our earlier work was followed. 4In a typical method, dried particles of pinecone biomass were grounded into finer particles in a ball mill.The dried biomass powder was then activated by carbonizing it in a muffle furnace at 600 ℃ for 2hrs at a heating rate of 20 ℃ min -1 by mixing it with KOH at a 1:2 ratio.The black carbonized sample was then cooled and washed with hot water and finally with cold distilled pH of the wash solution becomes neutral.The sample was then dried in an oven overnight at 110℃ and stored in an airtight sealed container, labelled as Pc-AC (Pinecone activated carbon) for further characterization and use.
The yield percentage of Pc-AC was calculated using; Yield ( %) = mass of final carbon after activation process initial mass of the dry impregnated sample × 100 Adsorbent's Characterization The prepared carbon was characterized through different analytical techniques.The carbon, hydrogen, and nitrogen percentage in the prepared carbon were determined using an elemental Analyzer (Model: PE 2400 Series II, Made: Perkin Elmer).The batch equilibrium method was employed to determine the zero point charge. 20To measure the pHzpc, a 50 mL solution of 0.1M KNO3 was prepared in a conical flask, to which the prepared Pc-AC was added.The adsorbent suspended was kept in this 0.01 M KNO3 solution until the pH reached a stable state.The initial pH values were adjusted in the range of 2 to 12 by adding HCl or NaOH.The mixture was continuously stirred for 2 days on a magnetic stirrer.Afterward, the final pH values were measured, and the difference between the initial and final pH values (∆pH = pHinitial -pHfinal) was plotted against the initial pH values (shown in Figure 1S in the supplementary information).Surface functionalities of the prepared carbon was determined using FT-IR Spectrometer (Model: Spectrum Two, Made: Perkin Elmer) in the range 4000 -400 cm -1 .Surface morphology was determined using scanning electron microscopy (JSM-6360, A c c e p t e d m a n u s c r i p t JEOL).The BET surface area and pore volume were measured by using a BET surface area analyzer (Smart instrument, SS93/02).

Preparation of standard fluoride stock solution
Standard fluoride solution was prepared by dissolving 2.178 g of NaF (Merck) in 1000 ml of double-distilled water in a 1000 ml volumetric flask.The solution was then further diluted to 100 mg L -1 of the fluoride solution.All test solutions of the desired concentrations were prepared by successive dilutions of 100 mg L -1 fluoride solution to get the required initial experimental concentration (2, 5 and 8 mg L -1 ).The pH of all the solutions was adjusted by adding 0.1N HCl or 0.1N NaOH before the addition of the adsorbent and was measured by pH meter (Make: Eutech).

Fluoride adsorption by batch mode experiments
For any adsorption study, a batch adsorption experiment is vital as it provides information about the different parameters that control the adsorption process.This method helps to find out the optimum condition for the reaction to be carried out effectively.Thus, adsorption parameters such as contact time, adsorbent dosage, pH, and fluoride concentration are investigated by this method.An optimized condition is established from the different experimental data and enables us to fix the experimentally obtained optimum condition to perform further experiments in the future.
In the current study, batch mode experiments were accomplished by taking 50 mL of fluoride known solution in an Erlenmeyer flask, which was stirred with a known weight of pinecone activated carbon at 150 rpm in a rotary shaker.The solution was filtered after the adsorption equilibrium was achieved, and the unadsorbed fluoride concentration was measured with the help of a Fluoride Meter (Model: Hanna Hi 98402 ISE Fluoride Meter).The effect of adsorption parameters like pH (pH 2 to pH 12), activated carbon quantity (0.05 to 0.30 g), contact time (0 min to 180 mins), and fluoride concentration of 2, 5, and 8  −1 were conducted.The removal percentage efficiency, and the defluoridation capacity ( −1 ) for Pc-AC was determined using equations 2 and 3 as given below: Defluoridation capacity (  ) = Where c0 = initial fluoride concentration in mg L -1 ; ce = equilibrium fluoride concentration in mg L -1 qe = amount of adsorbate adsorbed at equilibrium in mg g -1 ; M = mass of pinecone activated carbon in g; V = volume of fluoride solution in L.

Pinecone-activated carbon characterization
Results indicated that the yield and ash content values of Pc-AC were 16.88 % and 3.07 % respectively; the carbon, hydrogen, nitrogen, and oxygen content as determined by ultimate analysis was 87.69 %, 2.01 %, 0.72 %, and 6.51 % respectively (Table I).The BET surface area and the total pore volume of Pc-AC were 972.13 m 2 g −1 and 0.469 cm 3 g −1 as compared to the sample activated without any KOH which had a surface area of 108.54 m 2 g −1 and pore volume of 0.057 cm 3 A c c e p t e d m a n u s c r i p t g −1 .The higher BET surface area and pore volume upon KOH activation could be attributed to the elimination and release of volatile matter from the carbon skeleton to form voids, leading to an increase in the availability of active sites for adsorption.The pH of the zero point charge (pHzpc) of Pc-AC was found to be 6.93 reflecting the fact that the surface of the adsorbent was positive below this pH and negatively charged when pH > pHzpc. 21Figure 1(a) shows the SEM image of pinecone-activated carbon.The SEM analysis of the adsorbent indicates that the surface of the adsorbent was irregular and rough, with different pore sizes and shapes.The presence of such pore sizes is expected to influence the adsorption process and thus make it appropriate for the adsorption of fluoride.FT-IR study was used to determine the major surface functionality on the Pc-AC surface.The FTIR is shown in Figure 1(b).A very prominent peak around 3450 cm -1 was observed for Pc-AC which is due to the O-H stretching mode of hydroxyl groups, -NH group.It is also due to the adsorption and presence of moisture on the PC-AC surface.The functional groups include alcohols, phenols, carboxylic acids, and amine.The peaks around 1637 cm -1 account for the presence of C=O groups such as in carboxylic acids. 22While the band appearing at 1435 cm - 1 has been assigned to the aromatic structure of Pc-AC that are coupled to highly conjugated carbonyl groups or attributed to the stretching vibrations of C=O moieties of conjugated systems. 23Bands near 1395 cm -1 is assigned to C-O stretching vibration from ester, ether, and phenolic functional groups. 24While A c c e p t e d m a n u s c r i p t bands around 1048 cm -1 account for C-O vibrational stretching for primary and secondary alcohols.Finally, bands at 700-900 cm -1 corresponds to =C-H bending of substituted aromatic rings, while broad peaks at ~640 cm -1 account for O-H bending. 25The FT-IR studies show that various surface functionalities making them an ideal adsorbent for adsorption study.

Batch adsorption studies Effect of contact time and initial concentration
The effect of contact time on the adsorption of fluoride at three different initial concentration were investigated and reflected in Figure 2(a).Adsorption in all cases is found to increase rapidly with time and becomes saturated as time continues, and an equilibrium is achieved.The rapid adsorption in the initial period is because of the availability of active adsorption sites, but as time proceeds, these sites get accumulated with the adsorbate which does not allow any further adsorption.The equilibrium time for maximum fluoride adsorption was found to be 120 min for Pc-AC.
Moreover, it is well-known that adsorption depends on the relationship between unoccupied adsorption sites and the concentration of fluoride.Accordingly, figure 2(a) and supplementary figure 2S(a) shows that, with an increasing initial concentration of fluoride, the removal percentage decreases (increase in the adsorption capacity).This was due to the saturation of available adsorption sites on the adsorbent.It was observed that at 2 mg L -1 fluoride concentration, there seemed to be easy availability of unoccupied sites, which led to the complete adsorption of fluoride from the solution.However, as fluoride concentrations increased, a decrease in the number of active sites was noticed due to the increased adsorption of fluoride on the surface of the active sites which led to a reduction in available active sites.It was found that the removal percentage of fluoride by Pc-AC, at concentrations of 2, 5 and 8 mg L -1 were 99.9 %, 98.7 %, and 83 % respectively.Based on these observations, the fluoride concentration of 5 mg L -1 concentration was taken for all other experiments as the optimum concentration.

Effect of adsorbent dose
To confirm the proper dosage of the adsorbent, the adsorption process was conducted at different PC-AC dosages.Figure 2(b) and supplementary figure 2S(b) shows the effect of adsorbent dosage (0.05-0.30g) on the removal of fluoride by PC-AC.It was found that with an increase in the adsorbent dosage, the fluoride removal efficiency increased but reached equilibrium at an adsorbent dosage of 0.15 g with no further increase in the removal percentage after that.The high removal percentage at this dosage was due to the availability of active binding sites on the adsorbent as a result of its surface area. 26The porous nature of the adsorbent also facilitates fluoride adsorption, which is consistent with the observations obtained from the scanning electron microscopy (SEM) analysis.However, with a further increase in the adsorbent dose, there was no further increase in the removal rate (as there was a decrease in the adsorption capacity).This was due to accumulation or partial overlapping, which lead to a decrease in the total A c c e p t e d m a n u s c r i p t adsorption sites for fluoride. 27Thus, for subsequent studies, an adsorbent dose of 0.15 g was used.

Effect of solution pH
The pH of the solution is one of the basic parameters that affect the adsorption of fluoride.A study on the effect of pH on the adsorption of fluoride was done at different pH from 2 to 12 keeping other parameters like adsorbent dose and fluoride concentration constant (Fig. 3).It can be seen that; the adsorption of fluoride was more significant at lower pH (acidic medium) as compared to the alkaline condition.This could be because at lower pH (pH < pHzpc) most of the surface becomes positively charged which further facilitates the adsorption of fluoride.The maximum fluoride adsorption was found to take place at pH 4. But below pH 4, the fluoride removal was found to decrease due to the formation of weak hydrofluoric acid. 28On the other hand, when the solution pH was alkaline, there was a reduction in the adsorption of fluoride because of the competition between negatively charged surface hydroxyl ions and fluoride ions on active sites. 29hus, for further studies, the optimum fluoride concentration was taken at 5 mg L -1 with an adsorbent dose of 0.15 g at 120 minutes.As maximum fluoride removal takes place at pH 4, all further adsorption experiments were done at a pH of 4.

Isotherm studies
The study of adsorption isotherms is important as it elucidates the most significant isotherm model that determines the interlaying relationship between the adsorbate and adsorbent, useful for designing purposes.The experimental equilibrium data were modelled using three classical adsorption models: the Langmuir, Freundlich, and Temkin isotherms.These models were employed to describe fluoride adsorption on Pc-AC.The equations and the calculated equilibrium isotherm parameters are given in Table II.0.0174 qe = adsorption caPc-ACity at equilibrium (mgg -1 ); Ce = residual concentration at the equilibrium state of the system (mgL -1 ); Qmax = maximum adsorption caPc-ACity (mgg -1 ) KL = Langmuir constant; KF = Freundlich constant (L mg -1 ); RL = separation factor (provides insight into the favourability of the adsorption process); 1/nF = intensity of the adsorption, respectively; BT = Temkin constant related to the heat of sorption (Jmol -1 ); AT = equilibrium binding constant that corresponds to the maximum binding energy (Lg -1 ) Fitting of data with Langmuir isotherm for the adsorption of fluoride onto Pc-AC as shown in figure 3S(a), gave the qmax (maximum adsorption capacity) value of 2.845 mgg −1 .The correlation coefficient R 2 for Langmuir isotherm for fluoride adsorption by Pc-AC was found to be 0.999 (Table 2).Further, the separation (RL), was found to be 0.011, thereby indicating favourable adsorption of fluoride onto Pc-AC.The Freundlich isotherm, the plot of the model is shown in supplementary Figure 3S(b) and the parameters calculated from the slope and intercept are reflected in Table II.It can be seen that, 1/nF, for adsorption of fluoride by Pc-AC was found to be 0.22, thereby suggesting the favourability of fluoride adsorption onto Pc-AC.It also indicates that adsorption follows a normal Langmuir isotherm. 30The correlation coefficient R 2 value for fluoride adsorption by Pc-AC for Freundlich isotherm was found to be 0.906, which, when compared to Langmuir isotherm is way much lower.While fitting of the experimental data onto the Temkin isotherm model for fluoride adsorption by Pc-AC is shown in Figure 2S(c).The calculated parameters for the Temkin model are given in Table 3.5.The heat of adsorption (BT) was found to be 0.358 kJ mol -1 .The correlation value for Temkin isotherm for fluoride adsorbed onto PC-AC was 0.980.
For determining the appropriateness of the fit, the best adsorption isotherm model for the adsorption of fluoride by Pc-AC, Chi-square analysis was performed on all the isotherm models, and the results are shown in Table II.For adsorption of fluoride by Pc-AC, the Chi-square value for the Langmuir isotherm model was A c c e p t e d m a n u s c r i p t found to be the least and therefore, the most suitable to describe the adsorption of fluoride.Since the Langmuir equation assumes that the surface is homogeneous, it indicates a homogeneous distribution of adsorption sites on the adsorbent and the adsorption of fluoride by Pc-AC took place as monolayer coverage on the carbon surface containing a finite number of adsorption sites of uniform energy.The isotherm study also suggests that there is no transmigration of adsorbate in the plane of the adsorbed site.
The fitness of the isotherm for fluoride adsorption by Pc-AC follows the order: Langmuir > > Temkin >Freundlich

Adsorption Kinetics
The adsorption kinetics for the removal of fluoride by Pc-AC was investigated with pseudo first order, pseudo second order kinetic models and intraparticle diffusion model using the equation provided in Table III.The graph of log(qe−qt) versus t was plotted for pseudo-first-order while for pseudo second order kinetics models the graph of t/qt versus t was plotted and for intraparticle diffusion model, qt is plotted against t 1/2 .Figure 4S (provided in the supplementary information) shows the plot of pseudo first order kinetics, pseudo second order kinetics and intraparticle diffusion model for fluoride adsorption by Pc-AC.The results of the kinetic studies are listed in Table III.2.656 14.089 0.101 0.918 2.920 0.021 0.991 0.169 0.748 0.942 qe = the amounts of adsorbate (mgg −1 ) adsorbed at equilibrium; qt = amounts of adsorbate (mgg −1 ) adsorbed at a given time t (min); k1 = the rate constant of adsorption (min −1 ); k2 = pseudo-second-order adsorption rate constant (g/mg.min);kipd = intraparticle diffusion constant; C =intraparticle diffusion constant Table III shows that the measured qe value of the pseudo first order kinetic model do not tally with the experimental qe values for the adsorption of fluoride and even the appropriateness of the fit was not satisfactory.In the case of the pseudo-second-order kinetics model, measured equilibrium adsorption capacity, qe,meas of Pc-AC was found to be in good agreement with the experimental data.A good correlation coefficient (R 2 ) value of 0.999, 0.997, and 0.991 was found for pseudo second order kinetic model at 2 mL -1 , 5 mL -1 , and 8 mL -1 of fluoride A c c e p t e d m a n u s c r i p t concentration which in the case of pseudo-first-order model was 0.927, 0.930, and 0.918 respectively.Thus, the suitability of the pseudo second order kinetics model is seen to describe the adsorption of fluoride onto Pc-AC which suggests that the overall rate of adsorption is predominantly chemisorption.
It was established through SEM analysis that the prepared adsorbent is porous; thus, it is essential to consider how fluoride diffuses into the pores of Pc-AC.The experimental data were therefore, examined using the intra-particle diffusion model to gain insights into the diffusion of fluoride adsorption onto Pc-AC.Figure 4S(c) in supplementary information illustrates three distinct sections in the plot.The first straight portion corresponds to the rapid transport of the fluoride from the bulk liquid to the boundary layer surrounding the adsorbent due to the readily available adsorbing sites on the adsorbent surface.It represents fast adsorption of fluoride on the external surface of pinecone activated carbon via macropore diffusion.The second linear portion accounts for the transport of fluoride from the adsorbent surface into interior sites.The final step accounts for the slow diffusion of fluoride into micropores which are the less accessible sites of adsorption.It can also be seen that the magnitude of C increases with the increase in fluoride ion concentrations indicating an increase in boundary layer effects.
When compared with the pseudo-second-order kinetic model, the R 2 value for intra-particle diffusion was lower, as evident from figure 4S(c).Additionally, the intercept of the line does not pass through the origin, indicating that the adsorption mechanism of fluoride onto Pc-AC does not conform to this model.This observation adequately suggests that chemisorption predominantly controls the overall adsorption rate, following pseudo-second-order kinetics.[37][38][39]

Effect of co-ions on fluoride adsorption
The presence of co-ions like nitrate, sulphate, carbonate, chloride, and bromide normally occurs in fluoride-contaminated water.1][42] Thus, it is important to study the effect of co-ions to understand the efficacy of the prepared adsorbent.Therefore, the adsorption of fluoride in the presence of competing anions such as bromide (Br-), chloride (Cl -), nitrate (NO3 -), carbonate (CO3 2-) and sulphate (SO4 2-) was investigated.For this study, 20 mg L -1 concentration of anions was stirred with an initial fluoride concentration of 5 mg L -1 ; pH at 7; contact time for 120 minutes for PC-AC, at temperature 303 K and an adsorbent dose of 0.15 g.
As an effect of the presence of co-ions, the adsorption capacity was observed to decrease for carbonate, followed by sulphate, nitrate and chloride as depicted in Figure 4. We found that the defluoridation efficiency reduces significantly in the presence of sulphate and carbonates, i.e., for sulphate, adsorption of fluoride reduces from ~99 % (without any co-ions) to 84.17 %; for carbonate, adsorption  The reduced fluoride adsorption in the presence of sulphate could be due to the formation of both outer-sphere and inner-sphere surface complexes. 43On the other hand, the reduction due to the presence of carbonates could be because of the reduction of the positive charges on the active sites of the adsorbent by carbonates and thus reduces the active sites for adsorption of fluoride thereby lowering the removal efficiency. 44,45But, the presence of chloride and nitrate showed a lesser effect on fluoride adsorption as compared to sulfate and carbonate.This might be due to the outer-sphere surface complex formation (for chloride and nitrate).

Regeneration study
In an adsorption process, regeneration of the adsorbent is one of the crucial aspects as it reduces cost and determines the cost-effectiveness of wastewater treatment.A regeneration study also allows a researcher to understand the economic viability of the adsorbent.In this study, regeneration of the adsorbent saturated with fluoride was stirred in a 20 % NaOH solution for 1h.It was then washed with distilled water until the pH of the desorbed carbon became neutral and oven-dried at 110℃.For estimating the removal efficiency of the regenerated carbon, the adsorption experiments were repeated under the same conditions, i.e.,

Comparison of Pc-AC with other adsorbents
The fluoride adsorption efficiency for prepared activated carbon was compared with other adsorbents reported in the literature and is listed in supplementary Table S1.

CONCLUSION
The use of activated carbons derived from locally abundant pinecone biomass was investigated for fluoride removal from water in this study.The activated carbon exhibited an irregular surface and varied pores, as determined by SEM micrographs.FTIR analysis revealed the presence of different oxygen functionalities, including hydroxyl, ether, ester, carbonyls, etc. Batch adsorption experiments demonstrated that fluoride adsorption was favorable at a pH of 4 and followed a monolayer adsorption type, with a chemisorptive mechanism and a maximum adsorption capacity of 2.845 mgg -1 .The presence of co-ions, specifically sulfate and carbonate ions, influenced the fluoride adsorption process.The regenerated activated carbon using 20 % NaOH demonstrated promising results and could be used for fluoride removal for a certain cycle.In summary, the results indicate that pinecone-derived activated carbon exhibits promise as a sustainable and efficacious adsorbent for defluoridation purposes, particularly in

Fig. 2 .
Fig.2.(a) Effect of (a) contact time and initial concentration (b) Pc-AC dosage (c) pH on fluoride removal onto Pc-AC

A
c c e p t e d m a n u s c r i p t of fluoride reduces from ~99 % to 72.13 %.The reduction in the removal efficiency could be because of a change in the pH of the fluoride solution because of the addition of the competing ions (The pH values of fluoride solution -pH=7mixed with Cl−, NO3 − , SO4 2− , and CO3 2− increased for both the carbons as shown in Fig.3).So as the pH increases, the pH of the solution shifts towards the alkaline end, thereby reducing the fluoride adsorption efficiency.

Fig. 3 .
Fig.3.Effect of co-ions on fluoride removal by PC-AC

A
c c e p t e d m a n u s c r i p t 0.1 g of the regenerated samples with 50ml of 5  −1 of standard fluoride solution and agitated for 120 minutes at a pH of 7. It was found that the removal efficiency of the regenerated carbon was 93.23 % in the first cycle, 90.14 in the second cycle, 86.45 in the third, and subsequently decreased to 79.83 % in the fourth cycle as shown in Figure4.However, more studies are required in the future to increase the efficiency of the regenerated adsorbent in the subsequent cycles and also to explore the use of milder chemicals for the regeneration of the adsorbent.

TABLE I .
Properties of Pc-AC with and without KOH impregnation Yield ( %) BET Surface area (m 2 g −1 ) Pore volume (cm 3 g −1 )

TABLE II .
Different parameters of adsorption isotherm models for fluoride removal using Pc-AC

Table III .
Kinetic parameters for the adsorption of fluoride onto the pinecone-activated carbon