Cubical-Shaped Rods of Pectin-Hydroxyapatite Composite for Adsorption Studies of Fluoride by Statistical Method and Adsorption Experiments.

This research details the synthesis and application of a novel pectin-hydroxyapatite (PHAp) composite for fluoride (F-) adsorption from aqueous solutions. To determine the efficiency of the adsorption process parameters, i.e., adsorbent dose (0.1-0.4 g), initial fluoride concentration (10-30 mg/L), and temperature (298-313 K), the Box-Behnken design with three levels and three factors have been utilized. The quadratic model was established on 27 batch runs by regression analysis of the experimental data of these runs. The efficacy of adsorption was observed using the Langmuir and Freundlich models. The adsorption rate was found at 3.17 mg g-1min-1, and adsorption kinetics followed pseudo-second order (PSO) for PHAp. The significant novelty of this work is the synthesis of unique cubical-shaped rods biopolymer composite from hydroxyapatite. Additionally, this composite showed high adsorption capacity for F- compared to other hydroxyapatite adsorbents, and the improved adsorption capacity is attributed to its unique shape which provides a larger surface area. It can be reused for up to six cycles, which makes this method environment-friendly. The economic viability of the synthesized PHAp composite, in comparison to other adsorbents, is evident from the cost-benefit analysis.

Introduction

The fluoride (F–), released from industrial effluents, contaminates groundwater. This is of great concern as a high intake of F– is detrimental to human health, causing skeletal and dental fluorosis and neurological damages. Periodic weathering of rocks and minerals add F– into groundwater. The waste released from the glass and ceramic industries, electroplating, coal-fired power stations, and so on are other sources that increase the F– level in groundwater. These effluents can increase the F– level up to 10–1000 mg/L. The excess F– uptake through drinking water has an adverse effect on the health of several million people, primarily in developing countries. Although F– at low doses protects teeth from degradation, exposure to higher concentrations of fluoride can cause dental fluorosis. The beneficial dose and harmful levels are comparatively closer; the testified ideal value to avoid tooth decay is 0.5 mg/L, i.e., below the permissible limit of 1.5 mg/L suggested by the WHO. Hydroxyapatite (Ca10(PO4)6(OH)2, HAp), the key constituent of bones and teeth, is responsible for both detrimental and therapeutic health effects of fluoride. Due to the F– uptake nature of HAp, it has been used in water treatment.[1−8]

Therefore, it is necessary to design a highly selective and rapid method for elimination of F–. The conventional methods of F– removal are reverse osmosis,[9] nanofiltration,[10] ion exchange,[11,12] and adsorption.[13] Among these methods, adsorption is favorable due to its economic feasibility. Adsorption is suitable for removing contaminants in trace level and has become increasingly popular in water treatment applications owing to its simplicity, cost-effectiveness, efficiency, easy operation, and lower waste generation.

Recently, various low-cost adsorbents (i.e., activated alumina,[14] carbon nanotubes,[15] bone char,[16] fly ash,[17] metal oxide,[13] etc.) have been used for F– removal. The traditional adsorbents have limited application due to their low adsorption capacity; hence, research is vital to produce new and effective materials for F– removal. Of late, biosorption technique (an eco-friendly technique) has been predominantly used. Hydroxyapatite (HAp, Ca10(PO4)6(OH)2), the main essential mineral of vertebrate skeletal systems, is efficient as an adsorbent for removal of F–. HAp is a biomineral that is used in water treatment due to its eco-friendly nature, easy availability, low cost, and presence of excess exchangeable hydroxyl groups. However, their brittleness limits the application of HAp. Additionally, due to excess pressure drop during field application, its powder cannot be utilized in fixed-bed column directly. To overcome these technological issues, polymeric composites have been studied. Biopolymer-supported inorganic composites and their synthesis are in limelight these days due to their exceptional structure and properties.[18−20] Furthermore, the synergistic effect involved in these composites will provide additional mechanical strength to the composite.

Pectin is a type of bio-renewable, nontoxic, inexpensive, and biodegradable natural plant ionic polysaccharides, mainly composed of α-(1→4)-linked d-galacturonic acid.[21,22] These are biocompatible, biofunctional, nontoxic, nonimmunogenic, and biodegradable. Advanced studies mainly focus on biopolymer composites of HAp due to its high surface area and reactivity. In this study, a novel synthesis of pectin–hydroxyapatite (PHAp) composite with the utilization of new biopolymer in F– adsorption studies is discussed. Additionally, the unique feature of this composite is its shape, which enhances the adsorption sites as well surface area (157 m2/g), and hence the adsorption capacity (28.47 mg/g). Although to the best of the author’s knowledge, there is no work reported on differently shaped adsorbents for F– removal studies by using biopolymer composite adsorbent, Prabhu et al. reported dendrimer-like hyperbranched chitosan composite for F– removal having an adsorption capacity of 17.44 mg/g.[23]

The collection of statistical and mathematical techniques, which is called response surface methodology (RSM), was used for empirical model building to overcome the limitations of traditional methods. It is most convenient, particularly in adsorption or removal process,[24] for modeling mechanism parameters. The specific utilization of RSM is to dictate the most advantageous operational conditions for a specific system or to check a region that satisfies the operational specifications.[25,26] The successful synthesis of PHAp adsorbent has been done and it was characterized by Brunauer–Emmett–Teller (BET), field-emission scanning electron microscopy (FESEM), energy-dispersive X-ray analysis (EDAX), X-ray photoelectron spectroscopy (XPS), Fourier transform infrared (FTIR), X-ray diffraction (XRD), and thermogravimetric analysis (TGA) studies. Additionally, the effect of temperature, adsorbent dosage, initial concentration, and other factors on F– removal by Pec–HAp composite was inspected by using the Box–Behnken design (BBD) in RSM. Langmuir and Freundlich’s models were applied to study the adsorption mechanism, and pseudo-first-order (PFO) and pseudo-second-order (PSO) models were utilized to examine the kinetics of the adsorption process. The thermodynamic analysis was used to study the adsorption process feasibility, spontaneity, and nature.

The novelty of this research is the preparation of PHAp composite of unique shape (cubical rod shape), which has a high adsorption capacity due to the availability of six planes for the adsorption compared to other planar adsorbents, which provide only a single planar surface (Figure S1).

Discussion of Experimental Outcomes

Characterization

FESEM and EDAX Mapping

Figure shows the FESEM images of pectin (a, b), PHAp-F (c, d), and PHAp (e–h). The FESEM images clearly show the cubical rod shape of PHAp at different angles at 1–5 μm scale. Figure e shows the upper morphology of vertically standing rods, Figure f (5 μm) shows the agglomerated form of rods, and Figure g (5 μm) shows dispersed rod shape, which shows the dissimilar length of rods. After F– adsorption, the shape of the composite was destroyed, due to the accumulation of spherical shaped F– on the rods, which is clearly shown in Figure c,d. The pectin shows the irregular shape of the polymer. Moreover, Figure shows the EDAX spectrum of synthesized PHAp (Figure d), demonstrating the existence of C, O, Ca, and P elements, while in PHAp-F (Figure e), one sharp peak of the extra element appears in the EDAX spectrum. The EDAX spectrum of pectin shows only two peaks of C and O (Figure a), which shows the purity of the polymer as well as the synthesized compound.

Figure 1

FESEM images of (a, b) pectin, (c, d) PHAp-F, and (e–h) PHAp at different views of cubical rods.

Figure 2

(a, b, d) EDX images of pectin, PHAp, and PHAp-F and (c, e) mapping images of PHAp and PHAp-F.

Including the EDAX spectrum of the composite before and after F– adsorption, EDAX mapping images further supported the elemental composition of pectin, PHAp, and PHAp-F (Figure b,c,f, respectively).

BET

Figure A shows the N2 adsorption–desorption isotherms of type IV, and the corresponding pore size distributions are shown in Figure A1 of PHAp. The hysteresis loop (type IV) indicated the presence of both micropores and mesopores in the composite. The distributions of the pore volume, pore size, and the surface area before and after F– adsorption on PHAp are tabulated in Table S1. The BET surface area and the pore volume of PHAp and PHAp-F were found to be 157, 100.0 m2/g and 0.122, 0.981 cc/g, respectively, which were higher than the values reported in the literature.[27,28]

Figure 3

(A) Hysteresis curve, (A1) D–A pore size distribution, (B) FTIR spectra of PHAp and PHAp-F.

XPS Analysis

XPS was utilized to identify the various chemical species present on the adsorbent surface before and after the F– adsorption on PHAp. The high-resolution C 1s spectra (Figure A,B) were deconvoluted into four distinct peaks that correspond to the sp2–C, sp3–C, C–OH, and C=O bonds at 284.26, 284.50, 285.74, and 288.61 eV.

Figure 4

XPS scan of (A, B) carbon and (C, D) oxygen before and after F– adsorption, respectively.

Oxygen deconvolution spectra (O 1s) for PHAp shows three important contributions at 531.23, 529.37, and 528.83 eV (see Figure C,D). These signals can be associated with the oxygen interaction from chemisorbed water, calcium, and phosphorus in the PHAp composite. The intensities of the three signals are found to change after the interaction of F– with PHAp owing to surface modification, and these changes were observed at ∼528.56, 530.21, and 531.08 eV (Figure C,D). The decreased signal intensity was observed at 531.23 and 531.08 eV, which may be due to the interactions between the PHAp and F– ion. After F– adsorption, interactions of Ca with F–, i.e., binding energy of 347.88–351.49 eV, were observed, which also supported the presence of calcium on the surface of the adsorbent. After interaction with F–, their binding energy shifted to 348.07 and 351.63 eV (see Figure A,B). The P 2p XPS image exhibits two peaks at 132.9 eV assigned to P–C and P–N bonds, and the other peak at 133.26 eV corresponding to the P–O bonds of PHAp[29] in Figure C,D. The binding energies of all elements before and after F– adsorption are tabulated in Table S2.

Figure 5

XPS scan of (A, B) calcium and (C, D) phosphorous before and after F– adsorption, respectively, and (E) F–.

Finally, F 1s XPS images of F–-loaded PHAp are reported in Figure E, which contain two signals at 684.51 and 683.50 eV. The peak at 684.51 eV is an evidence of the surface reaction between F– and Ca2+, which implies the electrostatic interaction mechanism of adsorption. The second peak at 683.50 eV may be due to the replacement of −OH group of PHAp. The above results indicate the important role of hydroxyl groups in F– adsorption, implying the ion-exchange mechanism of F– adsorption.

FTIR Spectroscopy

Strong absorption bands appearing at 1024 and 621 cm–1 showed the stretching and bending vibrations of PO43– of PHAp (Figure B). The absorption band at 1024 cm–1 was broad and appeared due to the overlap of C–O–C stretching of pectin and PO43– stretching of PHAp. The absorption bands occurring at 1613 and 2935 cm–1 were attributed to the stretching vibrations of C–O and C–H groups, respectively. There are two broad bands observed between 2500 and 3550 cm–1. The broad band found in the region from 2500 to 3300 cm–1 is due to the stretching of O–H bond in the acid group, and another broad band appearing in the region from 3200 to 3550 cm–1 is attributed to the stretching of O–H band in alcohol group. The intensity of the broad band, appearing from 2500 to 3500 cm–1, of −OH bond in the acid group was decreased in the F–-adsorbed PHAp composite due to the exchange of the −OH group present in the composite by F–. The extra peak at 465 cm–1 is attributed to the F– peak.[30,31]

XRD

From Figure A, the crystalline peaks of HAp appearing at 2θ (25.92, 32, 33, 35, and 39.81°) were found in the PHAp composite, which showed that no obvious changes were found in the peak structure after the formation of composite and confirmed that the crystal structure of n-HAp was intact in PHAp composite.[32]

Figure 6

(A) XRD pattern and (B) TGA curve of PHAp.

TGA

The TGA curve of the PHAp (Huo et al.) exhibits the loss of mass in three steps. The first step, in general, is attributed to a weight loss of about 1.25 mg (10%) in the temperature range of 25–200 °C, which may be due to release of water within this temperature range, while the second degradation step, with 4% weight loss in the temperature range of 200–400 °C, was found to be due to the removal of −OH groups. The third degradation step contributed to a weight loss of 1.25 mg (10%) in the temperature range of 400–800 °C assigned by weak endothermic peaks could be related to some phase transformations[33,34] in structures (Figure B).

Development of Regression Model Equation for RSM

The RSM methodology of statics was utilized to construct a relation between the %A of F– and its depending parameters, i.e., three independent variables (adsorbent dose, Ci, and temperature), which directly affect the adsorption. The relation between independent variables and %A is in the form of a second-order polynomial equation. The polynomial equation, in the quadratic form, is designed on the analysis of coded or actual factor for %A of F–, and the equation obtained is shown in eq The positive or negative term defines the synergistic or antagonistic effect of the term, respectively. Analysis of variance (ANOVA) results of F– adsorption by PHAp are tabulated in Table . ANOVA was utilized to study the accuracy of the generated model, as well as to examine the fitness of the model and the main and interaction constants of the polynomial equation. The F value of the designed model was found to be 29.97, with lower probability (<0.0001) indicating that the model was significant.

Table 1

Response Surface Regression: %A vs Temperature, Adsorbent Dose, and Initial F– Concentration

parameters	coefficient	seq. SS	F	p
A	5.43	532.68	610.28	0.00
B	6.47	713.54	817.48	0.00
C	–1.98	70.84	81.16	0.00
AA	–1.28	9.92	11.36	0.007
BB	–2.65	42.26	48.41	0.00
CC	0.88	0.09	0.10	0.082
AB	–2.96	105.49	120.86	0.00
AC	–0.0005	0.00	0.00	0.99
BC	1.61	35.16	40.28	0.00
constant	88.39

From the polynomial eq , it can be concluded that of the three variables, B(Ci), initial F– concentration had the major effect on the %A of F–, due to the maximum F value. Variable A, i.e., adsorbent dose, is followed by B, again according to the F value. p > 0.05 is not significant for the designed model. In adsorption studies, all linear terms are significant, i.e., A, B, C (adsorbent dose, Ci, temperature), and quadratic terms (AA, BB) are significant, whereas only AB and BC are significant for the interaction terms. Other variables such as CC and AC do not have a significant effect on the F– adsorption due to the p-value which is >0.05.[31,35−37]

Effects of Variables on F– Adsorption

To understand the interaction between independent variables and their results, three-dimensional response surface plots (RSP) and two-dimensional counter plots of the designed model were constructed by utilizing the MINITAB 16.0 software (Figures –9 and S2). The RSP envision the interaction effects of each independent variable, which influence the %A of F–. The shape of the contour plot shows the nature and extent of the interactions effects among the experimental factors on the %A.[38] In each plot, the interacting variables were changed within the experimental assortments, while the third variable was constant or at the highest level.

Figure 7

Cube plot for %A.

Figure 9

Contour and surface plots of interaction between adsorbent dose and temperature.

Contour and surface plots of interaction between Ci and adsorbent dose.

Effect of Adsorbent Dose, Ci, and Temperature

Figures –9 show the collective effect of two factors alternatively, i.e., adsorbent dose and Ci, adsorbent dose and temperature, Ci and temperature, etc. The cube plot in Figure shows all three factors together in the form of their %A values with all three levels. Adsorbent dose and Ci both show a positive effect, and temperature shows a negative effect on %A of F–, i.e., with an increase in adsorbent dose and Ci, %A increases, while with an increase in temperature, %A decreases (Figure ). In case of adsorbent dose and Ci, %A increases because at a higher adsorbent dose, there is a number of active sites and with high Ci, there is a number of F– ions for adsorption. Hence, with the increase in the adsorbent dose (0.4 g) and Ci, the adsorption will be comparatively higher.[39]

In case of adsorbent dose and temperature (Figure ), it was concluded that F– adsorption decreases with increase in temperature and increases with increasing absorbent dose. This observation might be due to the fact that at constant Ci, an increase in adsorbent dose increases both surface area and the availability of active sites on PHAp molecules, which leads to enhanced F– adsorption. In case of temperature and Ci (Figure S2) on the F– adsorption onto PHAp at adsorbent dose 0.4 g, F– adsorption increases with increase in Ci because when the Ci increases, the active sites of the adsorbent will be surrounded by a greater number of F–. The increase in %A is very little with increasing temperature. This may be due to the increased number of binding sites with increasing temperature and hence the augmenting adsorption.

Adsorption Property

Effect of Initial pH

In the adsorption process, pH of the solution plays a significant role: it controls the adsorption of F– on the PHAp at the adsorbent interface. The pH effect was studied at a Ci of 10 mg/L, and pH ranges from 3 to 11 (Figure A). Percent adsorption of F– increases up to 7.0 (9.5 mg/g) and then decreases from 7 to 11 (8.6 mg/L). This change can be explained by the surface charge of PHAp in both alkaline and acidic mediums. It is well known that in acidic medium, i.e., pH < 7, the adsorbent surface is highly protonated and hence maximum F– is adsorbed in acidic medium, due to opposite charges of adsorbent and F–, and the opposite is true for alkaline media[40] (pH > 7).

Figure 10

(A) Effect of pH and (B) co-ions on F– adsorption.

Effect of Co-ions

The effect of co-ions on qe of PHAp was analyzed by comparing the fluoride qe with co-ions and without co-ions, for which initially fixed concentration of fluoride solution was analyzed by fluoride meter and then a similar concentration of co-anion, such as SO42–, with the fluoride solution was analyzed. Similarly, other co-anions, viz., Cl, NO3–, and HCO3– were also analyzed with fluoride solution. Analysis (Figure B) results shows the decrease in qe with co-anions because they compete with each other for the active sites on the adsorbent surface. The competition of anions for the active sites is closely related to the charge/radius (Z/r) ratio, ionic radii, and their concentration. The order of Z/r values is as follows: Cl– (1/0.181 nm) < NO3– (1/0.179 nm) < SO42– (2/0.230 nm). Ions with high Z/r values have high affinity with adsorbent and hence the multivalent anions with smaller radii have greater affinity than the monovalent anions. Therefore, SO42– competes more and reduces qe greater than the monovalent (NO3– and Cl–) anions. But this trend is not applicable in the case of OH– and HCO3–, and the effect of these two anions may be regulated by the ionic radii. Due to almost similar ionic radii of HCO3– (0.157 nm) ion with F– (0.133 nm) ion, it competes with fluoride strongly and decreases qe more than other ions. The decrease in qe is also due to the increase in solution pH in the presence of HCO3–.[41−43] The F– ion fits better into the crystal structure of apatite due to its smaller ionic radii and it can substitute OH– in HAp and form the more thermodynamically stable fluorapatite (Ca10(PO4)6F2). Therefore, the order of %A in the presence of coexisting anions is Cl– < NO3– < SO42– < HCO3–. However, qe of the PHAp adsorbent is much lower in the presence of HCO3– ion, i.e., 96.25%, owing to its greater interaction with an adsorbent, which reduces the active sites for F– adsorption.

Adsorption Isotherms

To compute qe of PHAp for F– adsorption, Langmuir (Figure A)[44] and Freundlich (Figure B)[45] isotherms have been studied at four different temperatures (298, 303, 308, and 313 K) Table . Langmuir isotherm model illustrated the monolayer adsorption on the homogeneous surface. Langmuir parameters are related to the maximum adsorption capacity (qm), and binding affinity of adsorbate to the adsorbent (KL). A high qe of 28.57 mg/g at 298 K accounts for good adsorption behavior of PHAp compared to other biopolymer composites of HAp for F–, as shown in Table . A dimensionless separation factor RL describes the feasibility of the adsorption. The value of RL < 1 (Table ) signifies the effectual interaction between PHAp and F–.

Figure 11

(A) Freundlich, (B) Langmuir isotherms, (C) PFO, and (D) PSO models at four different temperatures 298, 303, 308, and 313 K.

Table 2

Adsorption Parameters of Langmuir and Freundlich Isotherms

	Langmuir					Freundlich
	1/qe = 1/(qm × KL × Ce) + 1/qm					ln qe = ln KF + 1/n(ln Ce)
temp (K)	qm	1/qm	KL	RL	R2 (L)	R2 (F)	1/n	N	KF
298	28.57	0.035	0.86	0.103	0.95	0.96	0.55	1.81	94.63
303	26.31	0.038	0.82	0.107	0.95	0.94	0.52	1.92	88.23
308	25.64	0.039	0.84	0.105	0.94	0.94	0.50	2	83.93
313	25	0.040	1.03	0.08	0.96	0.92	0.48	2.08	76.70

Table 3

Comparison of qm of PHAp with Other Adsorbents

s. no	adsorbents	qm
1.	NaP–HAp nanocomposite[46]	12.69
2.	cellulose@HAp[43]	4.20
3.	cetyltrimethylammonium bromide-coated hydroxyapatite powder[47]	9.39
4	nanohydroxyapatite/chitosan composite[32]	1.56
5	alginate-encapsulated nanohydroxyapatite[30]	3.87
6.	HAp[31]	3.12
7.	PHAp (present study)	28.57

The Freundlich isotherm illustrates the multilayer adsorption of heterogeneous systems and assumes that different sites have several adsorption energies involved. KF and 1/n represent the qe and intensity of the adsorbent, respectively, in Table . It was found that the data fitted with the Langmuir isotherm model, with the highest R2 value of 0.96.

Adsorption Kinetics

The effect of contact time is a significant factor for the adsorption studies, i.e., kinetic studies. Figure S3 shows the effect of a change in time on F– adsorption process. It was experientially found that with increase in time, the F– adsorption increases gradually and reaches up to an equilibrium position after 30 min. Consequently, the contact time for F– adsorption process was considered as 30 min. To understand the kinetics of the adsorption process, PFO and PSO kinetic models were utilized to correlate the solid–liquid adsorption. Figure C,D shows the kinetics of the adsorption process, and the PSO model is the best fit according to the R2 analysis. These results predicted that the adsorption mechanism of PSO was predominant and that the physisorption process[48,49] controlled the adsorption because k2 value decreases with increase in temperature, as shown in Table .

Table 4

Kinetic Parameters of PFO and PSO Models Including Regression Coefficients

PFO				PSO
ln(qe – qt) = ln qe – k1 × t				t/qt = 1/(k2 × qe2) + t/qe
temperature (K)	k1	qe	R2	k2	ln k2	qe	R2
298	0.27	9.97	0.89	0.081	2.5	10	0.99
303	0.22	10.17	0.70	0.051	2.9	9.9	0.99
308	0.12	4.34	0.75	0.048	3.03	9.7	0.99
313	0.10	3.49	0.85	0.041	3.17	9.4	0.99

The activation energy (Ea), defined as the minimum amount of K.E. required for the adsorption to occur, gives an assessed energy barrier that the adsorbate must have to overcome for adsorption to occur. The Ea can be calculated by fitting the kinetic constant of the PSO model by Arrhenius equation at different temperatures[50] (eq ).The Ea value was found to be 32.6 kJ/mol (Figure B), the value of Ea decides the type of adsorption, and the result of Ea suggesting the adsorption of F– is related to particle diffusion controlled in this case.[51]

Figure 12

(A) Regeneration cycles, (B) activation energy curve by pseudo-second-order constant, (C) van’t Hoff plot, and (D) activation energy by sticking probability.

Thermodynamic Study

The adsorption process was much influenced by temperature. To examine the spontaneity, feasibility, and nature of the adsorption process, the thermodynamic data plays an important role, which can be obtained from eqs to 5.As shown in Figure , the values of ΔH and ΔS have obtained parameters of the curve ln KC versus 1/T (Figure C), i.e., from the slope and intercept, respectively. From Table S3, the positive value of ΔH and the negative value of ΔG confirmed the endothermic nature of the adsorption process and the feasibility and spontaneous nature of adsorption, respectively. The positive value of ΔS showed an increase in randomness during the adsorption of F– at the solid–solution interface.[52]

A new concept of sticking probability (S*) was explained using modified Arrhenius eqs and 7. S* is related to the surface coverage (θ). This is a measure of the potential of an adsorbent to hold the adsorbate on its surface.[53]Figure D shows the graph of surface coverage and 1/T, and the intercept and slope of the graph give ln S* and Ea/R, respectively. The value S* decides the nature of adsorption. The reported value of S* reveals the physisorption nature of adsorption (Table S4).

Reusability Studies

The regeneration of PHAp is a significant aspect to utilize adsorbent in practical application (Figure A). Table S5 shows the analysis data of repeated %A of adsorption–desorption cycles. After every cycle, the filtered adsorbent was washed with 0.01 M HCl/NaOH and then used for the next cycle. This repeated process was done for six cycles, and after six cycles, the %A was observed to be decreasing, probably due to the replacement of −OH group by F– ions. Nevertheless, the regeneration efficacy of the PHAp could be improved by treating the used adsorbent with NaOH/HCl.

Cost–Benefit Analysis

The cost–benefit analysis for the adsorbent is an important factor for its economic viability. The developed PHAp composite was synthesized by utilizing AR-grade chemicals. The cost–benefit analysis confirmed that the F– removal cost by PHAp is reasonably good compared to other traditional adsorbents. The cost of PHAp is higher than some traditional adsorbents,[54−58] as shown in Table ; however, PHAp had a high adsorption capacity of 28.57 mg/g, which even remained higher (60%) after six successive adsorption cycles. Hence, the synthesized PHAp composite shows good F– removal efficacy in water.

Table 5

Comparative Cost of Traditional Adsorbent and PAF

adsorbent	qm (mg/g)	cost of adsorbent (U.S. $/kg)
activated carbon	2.25	21.1
activated alumina	16.3	12.1
bone char	4.5	1.6
hydrous ferric oxide	13.2	10.42
PHAp	28.57	25.79

Adsorption Mechanism of F– on PHAp

Thermodynamic studies, spectroscopic investigation, adsorption studies, and kinetic studies all illustrate that the mechanism of adsorption of F– on PHAp is mainly due to ion exchange, electrostatic interaction, and also by physio-type adsorption (Figure ). From the XPS and FTIR investigations, the adsorbent has replaceable OH group, which participates in the ion-exchange mechanism. This is also proved by pH studies. At lower pH, the adsorbent surface acquires a positive charge, and hence it electrostatically interacts with negatively charged F–, while at high pH, the opposite phenomenon works, i.e., adsorbent acquires a negative charge. Additionally, the Ca2+ ions present in PHAp interact with F– ions by electrostatic attraction.

Figure 13

Expected mechanism of F– adsorption.

Conclusions

In this work, PHAp composite, an effective adsorbent for F– adsorption, was synthesized by a simple precipitation method, which produces a cubical-shaped rod-type structure. The PHAp composite was established by several characterizing techniques: BET, FESEM, EDAX, XPS, FTIR, XRD, and TGA. The RSM was combined by BBD to regulate the effect of three factors, i.e., adsorbent dosage, Ci, and temperature, on F– adsorption. The adsorption process follows PSO kinetics and follows Langmuir type of adsorption, i.e., monolayer adsorption. The qm for the PHAp adsorption process was found to be 28.57 mg/g. Thermodynamically, the process was feasible, spontaneous, and endothermic in nature, as revealed by the thermodynamic parameters, i.e., ΔG, ΔS, and ΔH. The overall study indicated that PHAp composite can be used as an efficient adsorbent for F– removal from drinking water and other water sources.

Experimental Section

Materials

Chemical reagents predominantly involved pectin, Ca(NO)3·4H2O, NH3, NaOH, HCl, NaNO3, NaNO3, NaCl, NaF, (NH4)2PO4, etc. All chemicals were of analytical grade and utilized without further purification. Deionized (DI) water was utilized during the preparation of adsorbent and adsorption experiments.

Preparation of PHAp Adsorbent

Pectin (2 g) was dissolved in 100 mL of DI water at room temperature, and the mixture was continuously stirred to obtain a homogeneous solution. Then, 20 mL of 1 M (NH4)3PO4 solution was added dropwise into the polymer solution within 15 min. The pH of the solution was maintained up to 10 using 25% NH3 solution. Then, 20 mL of 1.67 M calcium nitrate solution was added to the above solution for 30 min and stirred vigorously for 1 h, after which the solution was left for 24 h at RT. Then, the solution was filtered and washed with DI water to remove some impurities, and the precipitate obtained was PHAp. The precipitate was dried at 100 °C for 6 h and then crushed into a fine powder. The expected arrangement of HAp in pectin polymer is represented in Figure . PHAp was also synthesized by varying pH, ripening time, and concentration of calcium ions (Table S6), which are shown in Figure S4. The fine powder obtained was PHAp, which was further utilized for adsorption experiments.

Figure 14

Expected arrangement of HAp in pectin polymer.

General Characterization

The N2 adsorption–desorption isotherms, surface area, and pore size distribution of PHAp before and after F– adsorption were calculated by using BET analysis. Before the analysis, the samples were kept in vacuum at 300 °C for 4 h for outgassing to remove the volatile gases. To study the surface morphology and element detection, field emission scanning electron microscopy and EDAX analysis were carried out. The binding energy and surface composition of PHAp in both stages were analyzed by XPS analysis. Furthermore, to confirm the functional groups involved in the mechanism and the composition of the adsorbent, FTIR analysis was done using Bruker-FTIR. And to define the crystallography of the structure, XRD analysis was conducted. TGA analysis also was done to know about the thermal behavior of the PHAp using PerkinElmer TGA-4000.

Statistical Analysis of Adsorption Process

The RSM design was utilized to examine the F– adsorption on the PHAp. Statistical design for RSM comprises three levels (−1, 0, and +1). The RSM design methods were carried out with three independent variables (adsorbent dose, initial fluoride concentration (Ci), and temperature (Table )). Table shows the data obtained from the software after 27 runs per experimental design. Equation demonstrates the effect of variables in terms of linear and quadratic interactions. Then, the statistical calculation was done by using regression coefficients to generate dimensional and contour maps from the regression models. Figure shows the details (step by step) for optimizing F– adsorption. Minitab software was utilized for the analysis of the experimental data.

Table 6

Independent Variables and Their Coded Levels to RSM Design

independent variables/factors	units	–1	0	+1
adsorbent dose (A)	g	0.1	0.25	0.4
Ci (B)	mg/L	10	20	30
temperature (C)	K	298	308	313

Table 7

Batch Runs for RSM Experiment

runs	%A	A	B	C
1.	74	10	0.1	298
2.	85	10	0.25	298
3.	91.8	10	0.4	298
4.	85	20	0.1	298
5.	90.5	20	0.25	298
6.	93.9	20	0.4	298
7.	89.3	30	0.1	298
8.	93	30	0.25	298
9.	95.25	30	0.4	298
10.	69.5	10	0.1	313
11.	80.5	10	0.25	313
12.	87.3	10	0.4	313
13.	83.5	20	0.1	313
14.	89	20	0.25	313
15.	92.4	20	0.4	313
16.	89	30	0.1	313
17.	92.66	30	0.25	313
18.	94.93	30	0.4	313
19.	66	10	0.1	308
20.	77	10	0.25	308
21.	83.8	10	0.4	308
22.	82.25	20	0.1	308
23.	87.75	20	0.25	308
24.	91.15	20	0.4	308
25.	88.16	30	0.1	308
26.	91.83	30	0.25	308
27.	94.1	30	0.4	308

Figure 15

Flowchart for optimization of F– adsorption.

The following second-order polynomial equation was utilized to analyze the experimental outcomes of adsorptionwhere Y is the predicted response (%A) and βo, β, β, β, and X, X are the offset term, linear effect, first-order interaction, quadratic effect, and independent variables constants of the model, respectively. ANOVA was applied to study the adsorption behavior, i.e., %A, and independent variable effects (interaction effect), to find the optimum level, and to assess the statistical parameters by means of RSM.

Adsorption Experiments

All adsorption tests were carried out with Ci in the range of 5–100 mg/L, adsorbent dose of 0.1–0.4 g, contact time of 5–45 min, pH of 3–11, temperature of 298–313 K, assorted ion effect (Cl–, NO3–, SO42–, and HCO3–), and regeneration studies for F– adsorption onto PHAp. The pH value of the solution was regulated by using 0.1 mol/L HCl or NaOH. The adsorption kinetics of F– on the adsorbent was studied at four different temperatures (298, 303, 308, and 313 K) and a particular time interval (5–30 min) with the adsorbent dose of 0.4 g/L, Ci of 10 mg/L, and pH of 7.0. The adsorption isotherm trials were examined by altering the Ci from 10 to 100 mg/L at 298, 303, 308, and 313 K. All other conditions are the same as in kinetic studies. The residual F– concentration after each adsorption experiment was analyzed by F– equipped ion meter and calculated using eqs and 10.where %A and qe are % adsorption and adsorption capacity, respectively.