Oxidative Polymerization of Aniline on the Surface of Sisal Fibers (SFs) as Defluoridation Media for Groundwater

Chemical modification of sisal fibers via in situ oxidative polymerization of aniline was conducted to examine their defluoridation capacity for fluoride from drinking water. The effects of polyaniline modifications have shown significant changes on the chemical moieties and defluoridation capacity of sisal fibers (SFs). FTIR peaks at 1440 cm−1 and 1560 cm−1 revealed the presence of benzoid and quinoid structures together with sisal fiber (SF). Thermal profiles confirmed the enhancement of thermal stability of polyaniline-modified sisal fibers (PAniMSFs). SEM microstructure also proved the surface roughening of SFs as a result of polyaniline modifications. Optimal batch adsorption parameters (pH, contact time, adsorbent dose, and initial concentration) were found to be 5, 60 min, 1 g, and 10 mg/L, respectively. Adsorption kinetics proved that the removal of fluoride follows pseudo-second-order model (K2 = 0.18 g. (mg·min)−1), while the adsorption isotherm well described by the Langmuir and Freundlich model with an experimental adsorption capacity of 2.49 mg/g. Hence, modifications and improvements are required to reduce the amount of fluoride to a permissible level and enhance the longevity and activity of adsorbent materials.


Introduction
Fluoride is considered an essential micronutrient for the human body to prevent dental caries and mineralization of hard tissues [1].Beyond the optimal level recommended by WHO, 1.5 mg/L, fuoride becomes toxic causing bone disease and mottling of teeth [2].Many scholars have reported that continuous fuoride uptake damages the liver, kidney, and nervous system [3].More than 200 million worldwide, of which about 8 million people living in the Ethiopian Rift Valley rely on groundwater and boreholes contaminated with fuoride as high as 33 mg/L [4,5].
Volcanic rocks in the East African Rift Valley possess a large deposit of fuorapatite mineral (Ca 5 (PO 4 ) 3 F) compared to analogous rocks in diferent areas.Te Main Ethiopian Rift Valley is geologically unstable associated with rifting by hydrothermal energy and hot climate, which facilitates the solubility of natural rocks containing fuorapatite and other minerals [6].Consequently, defuoridation at a point of use is a must to provide the public with safe drinking water.
Adsorption, a physicochemical method, is preferentially selected over precipitation, osmosis, and coagulation to remove excess fuoride from drinking water because it is simple, low cost, and easily regenerable [7].Several adsorbent materials have been tested for similar purposes.Activated carbon possesses a high surface area which is considered as the most efcient adsorbent.However, production cost and regenerability restricted its comprehensive use [8].Although the Nalgonda project was applied in diferent parts of the world, it was not socially accepted and fnally abandoned [9].Diferent minerals were also tested to remove fuoride from drinking water.However, the generation of sludge restricted their popularity [10].
Nowadays, composites and resins attract the attention of scientists to develop defuoridation media for drinking water.Polyaniline is rich in amine and imine polyfunctionalities [11], which makes it ideal for ion exchange and ion pair interactions.It also uses the cheapest monomer and the easiest to synthesize [12].Diferent reports proved that polyaniline composites have shown appreciable defuoridation efciencies from drinking water [13].
However, the limited processability in aqueous environments and formation of aggregation hampers the versatility of polyaniline which in turn reduces the adsorption efciency.
Fibrous materials have emerged as substrates to alleviate the processability issues and formation of aggregation of sole polyaniline [14].Sisal (agave sisalana) fber is technically chosen as a substrate to carry out the active functionality of polyaniline, which is functionally rich in hydroxyl and carbonyls ready to make physicochemical attachments, mechanically fairly strong, and withstands the deterioration from saline water [15].Furthermore, chemical polymerization of aniline on fbrous materials enhanced the distribution of amine and imine multifunctionalities responsible to capture fuoride ions [16].
Terefore, this work presented the surface modifcation of sisal fbers via in situ oxidative polymerization of aniline to evaluate their defuoridation capacity from drinking water using a fuoride analyser electrode.

Experimental Procedures.
Matured sisal leaves were harvested from Shafat, 10 km to the south of Mekelle, Tigray, northern Ethiopia, and washed using running water.Sisal fbers were extracted using soil retting by burying matured sisal leaves approximately 40 cm deep on the ground for three weeks in a moisturized soil [17].Te extracted fbers were carefully washed with distilled water and then air-dried in a shade for about two weeks.Tese fbers were shredded into pieces using a paper shredder.Ten, these fbers were sieved based on their particle size by using the desired mesh numbers.

Polymerization of Aniline on Sisal Fibers (SFs).
Approximately, 10 g sisal fber (SF) was soaked in 1 M acidifed aniline.An equivalent amount of 1 M FeCl 3 .6H 2 O solution was poured into the mixture dropwise under continuous stirring [18].Te polymerization process was adjusted for 24 h at room temperature.After termination of the reaction, the polyaniline-modifed fbers were fltered and washed thoroughly with excess dilute HCl solution and distilled water, respectively.Finally, it was oven-dried at a lower temperature.

Characterization of Polyaniline-Modifed Sisal Fibers (PAniMSFs).
Screening of the chemical functionality of free sisal fbers (SFs) and polyaniline-modifed sisal fbers (PAniMSFs) was performed using Fourier-transform infrared (FTIR) spectroscopy (Bruker Vector 22).Tey were put to a thermal experiment on a simultaneous thermal analyser (DSC-TGA, SDT-Q600) from 25 °C to 600 °C to examine the weight loss as a function of temperature.External surfaces of SFs and PAniMSF were monitored using scanning electron microscopy (SEM) (JSM-IT300LV, JEOL, USA) at 500x magnifcation equipped with energydispersive X-ray (EDS) spectroscopy.
2.5.Preparation of Fluoride (F − ) Ion Solutions.A stock solution of 1000 mg/L fuoride ion was prepared by solubilizing 0.221 g NaF in 1 L distilled water.An intermediate solution of 100 mg/L was prepared by the dilution method, and other working solutions were prepared by further dilution to the desired concentrations (1,5,10,15, and 20 mg/ L) [19].Blank and control solutions were prepared and performed throughout the experimental process.

Optimization of Batch Adsorption Parameters
2.6.1.Efect of Solution pH.Approximately, 0.5 g PAniMSF was mixed with 200 mL of 10 mg/L fuoride solution by tuning the pH range from 1 to 10 by using 0.1 M HCl and 0.1 M NaOH solution for 50 min.Te pH at the point of zero charge (pH PZC ) was determined by using the pH drift method with minor modifcations, by placing 0.1 g adsorbent in 100 mL of 0.01 M NaNO 3 solution [20].

Efect of Initial Fluoride
Concentration.An optimum amount of PAniMSF (adsorbent dosage of 1 g) was added into 200 mL of (1, 3, 5, 8, 10, 12, and 15 mg/L fuoride solution, respectively) and stirred at pH 5 for a duration of 60 min at room temperature.After equilibration and achievement of optimum points for all parameters, the solution was fltered through the Whatman flter paper.Te residual fuoride solution was analysed by using a fuoride ion selective analyser (Hanna, HI552202, Romania).
Finally, the regeneration of adsorption materials was checked to test longevity and recyclability using dilute HCl solutions.All adsorption tests were conducted two times in triplicate measurements, and the reported amounts of fuoride ions are the mean of the experimental fndings.

Column Adsorption.
Adsorption capacity was also checked in the column experiment.A transparent glass column (internal diameter: 50 mm and length: 100 cm) was used for the continuous removal of fuoride from fuoride 2 International Journal of Analytical Chemistry simulated solution.Te bottom end of the column was sealed of with a sieve pore size less than 50 mm.Te column was flled with 100 mm size of the adsorbent material.Te known concentration of fuoride solution (10 mg/L) was continuously fed in the downward through the column.To investigate the efect of the fow rate, a series of experiments were conducted at four diferent fow rates (5, 10, 15, and 20 mL/min).Te experiment was equipped with a stopper valve to control the fow rate.Ten, the fltrate was collected for a fuoride analysis using the fuoride ion selective electrode.

Fourier-Transform Infrared (FTIR) Spectroscopy.
FTIR spectra of SF and PAniMSF are shown in Figure 1.Te FTIR spectra of SF and PAniMSF possess common absorption bands beyond 3300 cm −1 that belong to the O-H [21] and N-H [22] stretching vibrations derived from cellulose and aniline, respectively.To some extent, peak broadening was observed when the polymerization of aniline occurred together with sisal fbers (Figure 1(b)).A very weak band at 2900 cm −1 belongs to C-H stretching vibrations of alkyl hydrocarbons derived from the fber [23].Te FTIR peak which belongs to hydroxyl functionality has shifted to the lower band after adsorption of fuoride ion (Figure not included here).Tis can be justifed by the facilitation of fuoride adsorption via substitution of hydroxyl groups [24].Te peak at 1730 cm −1 is peculiar to the stretching vibration of carbonyl (C�O) functionality of hemicellulose and lignin of sisal fbers (Figure 1(a)) [25].However, there is a marked reduction of peak intensity during polymerization due to the interaction of N-H moieties with the existing system.Growth of polyaniline on SFs confrmed the appearance of two specifc peaks at 1440 and 1560 cm −1 which represent polyamino functionalities, benzoid (-NH-B-NH-) n and quinoid (-N � Q-N � ) n rings, respectively.Tese peaks signify the inclusion of polyaniline in its partially oxidized form [18,26], which are totally absent in sisal fbers.

Termogravimetry Analysis (TGA).
Termal analysis is an essential method used to investigate the weight loss of a material with temperature.Termal analysis (TA) of SF and PAniMSF is shown in Figures 2 A and B, respectively.Te frst degradation peak, which extends up to 100 °C, belongs to mass loss due to evaporation of chemically and physically adsorbed moisture [27].However, the mass loss is less for PAniMSF.A very sharp degradation profle which stretches up to 360 °C represents the decomposition of cellulose, hemicellulose, and pectin derived from sisal fbers [28].Te last stage beyond 450 °C inferred the degradation of polyaniline, which is absent in the sisal fber (line A).Introduction of polyaniline further proved the enhancement of thermal stability of sisal fber [25,29].

Surface Morphology (SEM) Coupled with Energy-Dispersive X-Ray Spectroscopy (EDS).
Morphological investigations of SF and PAniMSF were carried out by a scanning electron microscope (SEM) as presented in Figure 3. SEM micrographs proved the polymerization of aniline causes roughening the surface of the fber (Figure 3(b)).Te EDS spectra (Figure 4(a)) revealed strong peaks at CKα0.21 keV and Okα0.51 keV which represent carbon and oxygen derived from sisal fbers.Additional peaks at NKα0.39 keV and ClKα2.62 keV (Figure 4(b)) belong to nitrogen and chlorine, suggesting the inclusion of the oxidized form of polyaniline during oxidative polymerization [30].
Furthermore, a new signal also appeared at Fkα0.94 keV (Figure 4(c)) showing the existence of considerable content of fuoride ions adsorbed on the PAniMSF surface [31] after the fuoride adsorption experiment.Tis confrmed that the PAniMSF surface was successful for fuoride capture.Mass of elemental oxygen showed a reduction trend after polymerization and adsorption of fuoride occurred.Probably, this could be evidence that oxygen-bearing species are replaced by fuoride ions during the adsorption process [24].favoured at the acidic pH range.Protonation of nitrogen due to polymerization of aniline enhanced the electrostatic interactions between amine and imine surfaces with fuoride ions in aqueous solution.Furthermore, fuoride ions can easily displace a highly polarizable chloride ion which exists on the surface and interface of polyaniline as a counter anion [32].Te fuoride uptake capacity increased steadily (up to 2.21 mg/g) when the solution pH began to rise up from 1 to 4. In acidic media, both hydronium and chloride ions infuence the surface chemistry of adsorption boosting protonation and doping, respectively.Moreover, such species can support the formation of cationic radicals favouring fuoride exchange [33].Maximum defuoridation of fuoride was achieved at pH 5 (pH < pHpzc � 6.08) where the surface charge of the adsorbents remained positive [34].At extreme acidity, the amount of fuoride uptake was reduced due to the formation of weak fuoro acids and complex species (HF, HF 2 − , and F − ) which can afect the accessibility and mode of interaction of fuoride [32].

Batch Adsorption Study of
A further increase in solution pH (6 to 10) resulted in a sharp reduction of fuoride uptake (1.93 mg/g to 0.67 mg/g) due to the competition for active sites between hydroxyl groups and fuoride ions [33].Furthermore, a reduction in adsorption capacity beyond pHpzc (pHpzc � 6.08) is rationalized by electrostatic repulsion between the fuoride ion and the surface of the adsorbents which is dominantly charged negatively [35].
It is important to note that SFs removed a substantial amount of fuoride although the removal capacity was not 0.00 1.00 2.00 3.00 4.00 5.00 6.00 1.00 2.00 3.00 4.00 5.00 6.00 1.00 2.00 3.00 4.00 5.00 6.00 Energy (keV)   International Journal of Analytical Chemistry comparable to polyaniline-modifed ones.Te hydroxyl functionalities of sisal fbers can interact with hydronium ions in the solution to form positively charged surfaces leading to fuoride uptake.Moreover, other weak van der Waals interactions may occur between fuoride and the sisal surface [36].Diferent scholars have applied cellulose, chitin, and chitosan as adsorption media due to the availability of hydroxyl functionality through H-bonding and other van der Waals interactions.However, their adsorption capacities were minimal in which modifcation and activation to acquire amines, carbonyl, and hydroxyl in their anatomical chains enhanced removal efciency [31,37].

Efect of Contact Time and Adsorption Kinetics.
Te efect of contact time on the removal of fuoride ion using PAniMSF is presented in Figure 6.Findings of adsorption capacity confrmed that the removal of fuoride was fast enough up to the frst 45 min.Approximately, 2.3 mg/g of the fuoride ion was defuoridated in a short time.Tis may be justifed by the availability of excess active sites and the high concentration gradient at the initial stage.It can be visualized (Figure 6) that an equilibrium was achieved beyond 60 min.Once equilibrium was achieved, the rate of defuoridation remained almost the same throughout the process due to further reduction of the number of vacant sites and the amount of fuoride ions in the solution [38].Sisal fbers also followed similar adsorption trends.However, the adsorption capacity is much lower than polyaniline modifcation.
In order to validate the experimental data and forecast the rate of fuoride uptake on the surface and interface of PAniMSF, Lagergren pseudo-frst-order [39] and pseudosecond-order [40] models are applied.Te linear versions of the Lagergren pseudo-frst-and second-order expressions are shown in equations ( 1) and (2), respectively: ln q e − q t  � ln q e − K 1 t, (1) where q e and q t (mg/g) are the adsorption capacities at equilibrium and time (t, min), respectively, K 1 (1/min) and K 2 (g/(mg•min)) are the rate constants for the Lagergren pseudo-frst-order and pseudo-second-order models, respectively, C o and C e are the initial and residual concentrations of fuoride (mg/L) at equilibrium, respectively, m is the mass of adsorbent (g), and V is the total volume of a solution (L).Rate constant (K 1 ) and adsorption capacity (q e ) at equilibrium are computed the slope and intercept of the plot of ln (q e − q t ) against time (t) from the linearized equation of the Lagergren pseudo-frst-order model (Figure 7(a)).Te graph is largely deviated from a straight line (regression coefcient, R 2 � 0.91).Te calculated adsorption capacity (q e , cal � 9.14 mg/g) determined from the graph (Table 1) difers signifcantly from the experimental adsorption capacity (q e , exp � 2.49 mg/g) (calculated using equation ( 3)).Tese results demonstrated that the Lagergren pseudo-frst-order kinetics is not appropriate for fuoride adsorption on PAniMSF.
Rate constant (K 2 ) and adsorption capacity (q e ) are calculated from the intercept and slope of the plot of t/q t against time t from the linearized equation of pseudo-second-order kinetics (Figure 7(b)).Te graph is satisfactorily ftted to a straight line (regression coefcient, R 2 � 0.98).Te calculated adsorption capacity (q e , cal � 2.65 mg/g) is approximately equal to the experimentally determined adsorption capacity (q e , exp � 2.49 mg/g).Tese results suggest that pseudo-second-order mechanism is favourable for the adsorption of fuoride.Defuoridation onto PAniMSF occurring for a short period may prove the existence of strong ionic interactions between the positively charged polyaniline surface and fuoride ions, referred as chemical adsorption [41].Rate constants (K 1 and K 2 ) and adsorption capacity (q e ) calculated from the plots of Lagergren pseudo-frst-and second-order kinetics are shown in Table 1.

Efect of Adsorbent Dosage.
To achieve greater adsorption capacity, an adsorbent dose of 0.25 g to 6 g was used to defuoridate 10 mg/L of fuoride solution (Figure 8).Fluoride uptake capacity was enhanced with a lessening in adsorbent dosage due to the accessibility of a higher number of fuoride species per unit mass of adsorbents [42].Te ability to adsorb fuoride was decreased from 2.32 mg/g (solid/liquid ratio � 1.25 g/L) to 1.23 mg/g (solid/liquid ratio � 5 g/L).Figure 8 illustrates that a further increase in the adsorbent dose (beyond 3 g) has little efect on the defuoridation capacity due to the very low equilibrium concentration of fuoride [43].Sisal fbers, on the other hand, only managed to reach a maximum of 0.97 mg/g during the trial, and when the dose of adsorbent was increased, they further decreased to 0.25 mg/g.

Efect of Initial Concentration and Adsorption
Isotherms.Results of adsorption capacity for initial fuoride concentrations are shown in Table 2. Fluoride take-up has ceaselessly increased with an increment in the sum of fuoride, whereas the amount of adsorbent remained unchanged and solidness accomplished.Tis can be legitimized by the presence of a higher number of fuoride than the number of adsorbents which consequently goes to saturation [44].However, polyaniline modifcation boosted the adsorption capacity (q e ) of sisal fbers (PAniMSF: 2.48 mg/g versus 0.82 mg/g of SF) which confrmed the inclusion of polyaniline moieties enhanced the surface chemistry of sisal fbers.
Langmuir and Freundlich isotherm models are frequently utilized to characterize the adsorption behavior of adsorbents in the liquid phase.Te Langmuir isotherm is utilized to measure the arrangement of monolayer adsorption on homogeneous adsorbent sites [45] with the thought that all adsorption destinations are indistinguishable and there is no interaction among adsorbed species on adjoining binding sites.Te Freundlich isotherm considers Table 1: Lagergren pseudo-frst-and pseudo-second-order kinetic parameters of fuoride ion adsorption at room temperature.

C o (mg/L)
Lagergren pseudo-frst-order q e , exp (mg/g)     International Journal of Analytical Chemistry adsorption taken at heterogeneous adsorption sites leading to multilayer formation [46].Te linearized forms of Langmuir and Freundlich isotherm models appear in equations ( 4) and ( 5), respectively: where q m (mg/g) and K L (L/mg) are Langmuir constants which belong to the maximum monolayer adsorption capacity and energy of adsorption, respectively, and K F (mg/g) and n F represent the multilayer adsorption capacity and the degree of dependence of adsorption at equilibrium concentration, respectively.Te empirical values of Langmuir adsorption constants (q m and K L ) are computed from the slope and intercept of the linear plot of C e /q e against C e , respectively.Essentially, Freundlich isotherm parameters (K F and n F ) are calculated from the intercept and slope of the plot of log q e versus log C e relationship.Te isotherm data are assessed by utilizing Langmuir and Freundlich isotherm models, and the corresponding results are shown in Figures 9(a) and 9(b), respectively.Te isotherm parameters calculated from the slope and intercept of the respective linear plots are given in Table 3.
Te regression coefcients of both Langmuir and Freundlich isotherms have palatably ftted to the straight lines for adsorption of fuoride ions in spite of the fact that minor variations exist.Te highest adsorption capacity (q m � 3.1 mg/g) calculated from the Langmuir isotherm graph showed an approximate value with the experimental adsorption capacity (q e , exp � 2.49 mg/g).Values of the separation factor (R L ) (computed utilizing equation ( 6)) are all found within the range 0 < R L < 1 [47] which suggests the favourability of the equilibrium state.Te adsorption intensity of the Freundlich isotherm lies within the favourable range 0 < 1/nF < 1 [48].Hence, both isotherm models ftted well for the adsorption of fuoride which includes physicochemical intuitive on the adsorbent surfaces.SEM microstructures (Figure 3) further demonstrated the occurrence of a nonuniform adsorbent surface, particularly when the adsorption cycles repeated:

Comparison of Adsorption Capacities of PAniMSF with Other Adsorbents. Comparison of adsorption capacity of
PAniMSF with other adsorption materials is troublesome due to the variety of experimental conditions.However, their execution can be generally assessed using adsorption capacity.Tus, the values of adsorption capacity of PAn-iMSF have appeared as average performance compared to the literature values (Table 4).Moreover, defuoridation includes either an ion-exchange mechanism or an electrostatic mode of interaction or both at a time to expel fuoride from water [50].However, data on the nature of adsorbents (such as particle size, porosity, and dose), solution pH, and dosage of pollutants are either missed or totally distinctive in context.

Regeneration Studies.
Adsorption is fnancially feasible.Consequently, any adsorbent material ought to be efortlessly recovered and reused as numerous times as conceivable [35].Recovery tests were carried out using diverse concentrations of HCl solution as a desorbing agent.Te adsorption test was conducted on the recovered adsorbent materials up to the seventh cycle with a small reduction in the adsorption performance.However, adsorption performance of the recovered adsorbents began to decrease essentially after the sixth adsorption cycle.

Column Study.
As adsorption is a time subordinate process, the impact of the fow rate on the adsorption of fuoride was explored by changing the fow rate from 5 to 20 mL/min, pH 5, and a dose of 1 g, and the corresponding results are shown in Figure 10.Adsorption capacity has shown a lessening pattern with an increasing fow rate due to the reduction in contact time between fuoride particles and adsorbents.However, lower fow rates give the chance to create superior adsorption contact, and thus, the defuoridation process is brought to an end.Defuoridation capacity of SFs (used as control) was substantially lower than PAniMSF.Adsorption capacity of sisal fbers (0.97 mg/g for 10 mg/L) is lower than the reported value of chitin (3.4 mg/g for 5 mg/L) [51] although the fow rate and initial fuoride concentration are diferent.

Conclusions
Tis study presents the development of defuoridation media from drinking water using sisal fbers modifed by polyaniline via in situ oxidative polymerization of aniline.FTIR, TGA, and SEM-EDX studies revealed that sisal fbers are chemically modifed using polyaniline.FTIR peaks at 1440 and 1560 cm −1 confrmed benzoid (-NH-B-NH) n and quinoid (-N � Q-N � ) n amino functionalities of polyaniline exist together with the surface and interface of sisal fbers.Detection of particular peaks at NKα0.39 keV and ClKα2.62 keV from EDX spectra belongs to nitrogen and chlorine which proved the introduction of polyaniline in the partially oxidized and doped state.Last, the fuoride uptake capacity was evaluated in batch and column experiments after optimization of adsorption parameters.Te corresponding experimental results proved polyaniline modifcation showed appreciable adsorption capacity, 2.49 mg/g.Adsorption removal of fuoride ions followed pseudo-second-order kinetics, while both isotherms Langmuir and Freundlich ftted best to the experimental data.
To sum up, PAniMSFs have the potential to remove fuoride from drinking water when the adsorption parameters are fully optimized and monitored.

8
International Journal of Analytical Chemistry

Figure 6 :
Figure 6: Efect of contact time on defuoridation of water (C o : 10 mg/L, pH: 5, and adsorbent dose: 0.5 g) at room temperature.

Figure 8 :
Figure 8: Efect of adsorbent dose on defuoridation of water (C o : 10 mg/L, pH: 5, and contact time: 60 min) at room temperature.

Figure 10 :
Figure 10: Te efect of the fow rate on defuoridation of water (C o : 10 mg/L, pH: 5, and adsorbent dose: 1 g) at room temperature.

Table 3 :
Langmuir and Freundlich isotherm parameters of fuoride ion adsorption at room temperature.

Table 4 :
Summary of fuoride uptake capacity of polyaniline-based adsorbents.