Cellulose bridged carbonate hydroxyapatite nanoparticles as novel adsorbents for efficient Cr(VI) removal

Abstract Removal of Cr(VI) from aquatic environment is crucial due to its bioaccumulation, high mobility and strong toxicity. In this work, a novel nano-adsorbent consisting of carbonate hydroxyapatite (CHAP) and carboxymethyl cellulose (CMC) was designed and successfully synthesized by a simple route for the uptake of Cr(VI). The synthetic CMC bridged CHAP (CMC-CHAP) material exhibited higher surface area (122.90 m2/g) and adsorption capacity for Cr(VI) (13.45 mg/g) than other apatite based adsorbents. The adsorption process of Cr(VI) by CMC-CHAP was in line with the Langmuir isotherm model and pseudo-second-order kinetic model. The calculated thermodynamic parameters showed that the adsorption of Cr(VI) on CMC-CHAP was a spontaneously endothermic process. In addition, CMC-CHAP had good ability to remove Cr(VI) under the interference of coexisting ions, and possessed remarkable reusability. Based on the pH-effect experiment and x-ray photoelectron spectroscopy characterization, the removal of Cr(VI) by CMC-CHAP was considered to be synergistic processes of electrostatic attraction, reduction reaction and chelation. This work provided new insights into performance optimization and application potential of CMC-CHAP on Cr(VI) removal from water. Graphical abstract


Introduction
Hexavalent chromium (Cr(VI)) is highly toxic, carcinogenic, and is often found in the effluent wastewater from various industries such as textile dyeing, chromite ore processing, electroplating, leather tanning, and battery manufacturing processes. [1,2]Due to easy migration, strong solubility, and high toxicity of Cr(VI), it can easily transfer into the aquatic organisms and ecosystems through the food chain and cell membranes. [3]Cr(VI) pollution need to be eliminated to alleviate the harmful effects on human health and environmental ecosystem.The United States Environmental Protection Agency has recommended that the maximum permissible limit of Cr(VI) in drinking water is 0.05 mg/L, while the limit for total Cr is 0.10 mg/L. [4,5]Therefore, the removal of toxic Cr(VI) from wastewater has become an urgent necessity.
To meet strict environmental regulations, numerous physical and chemical methods are being used to remove Cr(VI), such as solvent extraction, [6] reduction-precipitation, [7] membrane separation, [8] biological treatment, [9] electrochemical process, [10] and adsorption. [11]However, the majority of these methods have some drawbacks such as high operating cost, low efficiency and generate secondary pollution. [12]In comparison, adsorption is considered to be the most versatile and widely used method among these conventional techniques, due to its efficient selectivity and operational convenience. [11,12]Accordingly, it is particularly important to develop novel and effective adsorbents for water environment security.
Recently, the application of nano-adsorbents has emerged as a fast-developing, fascinating area of interest for the removal of Cr(VI) from water because of the distinct characteristics of nanomaterials.Nanosized carbonate substituted hydroxyapatite (CHAP, Ca 10 (PO4) 6-x (CO 3 ) x (OH) 2-y (CO 3 ) y ), which is a well-known member of the apatite family and the major inorganic component of natural bone, [13,14] have drawn tremendous attentions in eliminating various contaminants from water.The CHAP adsorbent has some key advantages over its conventional counterparts.First, CHAP adsorbent can be easily synthesized via a simple environmental-friendly process of neutralization/precipitation reaction without organic solvents. [15]Second, it is a suitable adsorbent for heavy metals and radionuclides due to its low water solubility, high stability under reducing and oxidizing conditions, and good buffering properties. [16]Third, both the chemically synthesized and biogenic CHAP materials are inexpensive as compared to the commercial adsorbents, such as activated carbon. [17]Last but not least, nanosized CHAP exhibits excellent biocompatibility and adsorption properties, and has been considered as an environmental benign functional material. [18][21][22][23] One reason for the unsatisfying Cr(VI) removal performance by CHAP lies in the fact that the fact that Cr(VI) exists in water as oxyanions like Cr 2 O 7 2-, HCrO 4 -and CrO 4 2-, which do not form insoluble compounds in aqueous solution. [24][24][25] Furthermore, the CHAP nanoparticles tend to rapidly agglomerate to form larger aggregates due to the van der Waals forces, causing loss in dispersibility and reduced reactivity.
To address this issue, stabilization strategy was commonly adopted using water-soluble polymer (such as starch or carboxymethyl cellulose (CMC)), which could attach on the nanoparticles and prevented them from aggregating thorough the interparticle electrostatic and/or steric repulsions. [26]For instance, Li et al. [27,28] utilized CMC as a bridging agent to coat partially on chlorine substituted apatite nanoparticles through bidentate bridging and hydrogen bonding, which made the apatite nanoparticles not only maintain their high surface area and adsorption ability to metal(oid)s (Cd 2þ , Zn 2þ and As(V)) but also can be easily separated via flocculation and gravity settling.As inspired by the literature, it was expected that the CMC bridged CHAP nanoparticles can offer both high adsorption/reduction ability for Cr(VI) oxyanions and easy separation nature.However, no relevant report has so far been focused on Cr(VI) removal by CHAP nanoparticles bridged with CMC.Also, the underlying mechanism of Cr(VI) removal by this material still has not been explored.
In this work, the physical and chemical properties of CMC bridged CHAP nanoparticles and their adsorption potential for Cr(VI) were investigated by comprehensive physicochemical and intermittent adsorption experiments.The overall objective of this study was to: (1) synthesize and characterize CMC bridged CHAP nanoparticles at various CHAP to CMC mass ratio; (2) determine the effects of initial pH, contact time, initial Cr(VI) concentrations, and coexisting anions on Cr(VI) removal efficiency; and (3) acquire further insights into the underlying Cr(VI) removal mechanisms by CMC bridged CHAP nanoparticles.

Materials
In this study, Ca 2 (NO 3 ) 2 �4H 2 O, (NH 4 ) 2 HPO 4 , (NH 4 ) 2 CO 3 and ammonium hydroxide were used as starting raw materials.The CMC (M.W. ¼ 90,000 in sodium form, degree of substitute ¼ 0.7) was provided by Aladdin Industrial Corporation (Shanghai, China) and used as the bridging agent.While other chemicals were purchased from Nanjing Chemical Reagent Co., Ltd.(Nanjing, China).Deionized water was applied in the whole experiment process.All chemical reagents were of analytical grade and were used without further purification.

Preparation of CMC bridged CHAP
For the synthesis of CMC bridged CHAP (CMC-CHAP) adsorbent, certain concentrations of CMC, Ca 2þ and PO 4 3solutions were prepared separately using NaCMC, (NH 4 ) 2 HPO 4 and Ca(NO 3 ) 2 in deionized water.The preparation of CHAP was based on a chemical precipitation method, which has been described in detail in our previous work. [15]Typically, the Ca 2þ (0.25 M, 150 mL) was dropwise added to the CMC solution (300 mL, 0, 0.25, 0.5 and 1.0 wt %) and stirred for 12 h to yield the Ca solution.The P solution was prepared by mixing appropriate amount of (NH 4 ) 2 HPO 4 and (NH 4 ) 2 CO 3 to obtain a 150 mL PO 4 3solution containing 0.15 M PO 4 3-and 1.0 mM CO 3 2-. [15]The aforementioned P solution was dropwise added to the Ca solution with constantly stirring.The pH value of the mixture was adjusted by ammonia solution to 11 during the addition process.To ensure complete reaction and full growth of the CMC-CHAP particles, the suspension was kept at room temperature in a sealed beaker for 12 h with vigorous stirring before the precipitates were collected by centrifugation.The precipitates were washed with ethanol thoroughly to remove residual cations and anions, and then dried in air at 65 � C for further characterization.The obtained samples were labeled by the CMC contents.For example, 0.25CMC-CHAP, 0.5CMC-CHAP, and 1.0CMC-CHAP, represented the CHAP samples prepared in the presence of 0, 0.25%, 0.5% and 1.0% CMC, respectively.For comparison, bare CHAP particles were also prepared without CMC but under otherwise identical conditions.

Characterization
X-ray diffraction (XRD, Bruker D8 Advance, Bruker Corporation, Billerica, MA, USA) measurement was performed on a diffractometer with Cu Ka radiation with k ¼ 1.5406 nm Å.The Fourier transform infrared (FTIR) spectroscopic analysis using KBr pellets was performed using a Nicolet iS5 (Nicolet Instrument, Thermo Company, USA) spectrophotometer.The specific surface area of the samples was measured by the N 2 adsorption technique, using an ASAP 2460 (Micromeritics Instrument Corporation, Norcross (Atlanta), GA, USA) analyzer based on the Brunauer-Emmett-Teller (BET) method.Scanning electron microscopy (SEM) images were obtained using a JSM-6700F (JEOL, Ltd, Tokyo, Japan) microscope.The elemental composition was determined by energy-dispersive x-ray spectroscopy (EDS, attached to the SEM).Transmission electron microscopy (TEM, Titan G2 60-300, FEI Company, Hillsboro, OR, USA) technique was used to visualize the morphology of the prepared samples.After Cr(VI) adsorption, the chemical states of the surface elements were recorded by x-ray photoelectron spectrometer (XPS) using an ESCALAB 250Xi (Thermo Fisher Scientific, USA) spectrometer equipped with an Al anode (Al-Ka ¼ 1486.7 eV) as x-ray source.

Adsorption experiments
Batch adsorption experiments were conducted by transferring 50 mL of Cr(VI) solution into 100 mL of the conical flasks at 298 K, using an identified quantity of CMC-CHAP as adsorbents.Then the conical flasks were agitated at 180 rpm using a magnetic stirrer, allowing sufficient time for adsorption equilibrium.The solution pH was adjusted using 0.1 M HNO 3 or 0.1 M NaOH.After reaching adsorption equilibrium, the solid materials and adsorbate were separated by a TDL-500C centrifuge (Shandong Baiou Medical Technology Co., Ltd., Jinan, China) at 5500 rpm for 15 min.The supernatants were filtered through a 0.22 lm syringe filter prior to assaying the Cr(VI) concentration, which was analyzed by 1, 5-diphenylcarbazide method with Metash UV-8000 (Shanghai Yuanxi Instruments Co., Ltd., China) spectrophotometer at 540 nm. [24]The adsorption capacity (mg/g) was calculated as follows: q e ¼ (C 0 -C e )V/m, where C 0 and C e are the initial and equilibrium concentrations of Cr(VI) in solution; V (L) is the volume of adsorbate solution; and m (g) is the weight of adsorbent.All the experiments were carried out in triplicate and mean values are reported.
Batch sorption kinetic tests were carried out in 100 mL conical flasks.The initial Cr(VI) concentrations were 5, 10 and 20 mg/L, respectively.The dosage of CMC-CHAP was 0.1 g/50 mL.The pH value was adjusted to 3 with 0.1 M HNO 3 or 0.1 M NaOH.The flasks were then agitated at 180 rpm at 298 K.The remaining Cr(VI) concentration at different time intervals was measured to calculate the adsorption capacity (q t ).The pseudo-first-order (PFO), pseudo-second-order (PSO) and intraparticle diffusion models (described in the Supplementary material) were adopted for kinetic fitting.
In the adsorption isotherm experiment, 0.1 g adsorbent was placed in 50 mL Cr(VI) solution with contents of 5, 10, 15, 20, 25 and 30 mg/L at 288, 298 and 308 K to agitate for 6 h.Then, it reached adsorption equilibrium and calculated the equilibrium adsorption capacity q e (mg/g).The Langmuir, Freundlish and Sips models ((described in the Supplementary material)) were applied to fit the adsorption isotherm data.
Effect of coexisting ions on Cr(VI) adsorption was conducted by shaking 0.1 g of CMC-CHAP adsorbent in 50 mL of 10 mg/L Cr(VI) solution containing K þ , Na þ , Ca 2þ , Mg 2þ , Cl -, NO 3 -, or SO 4 2-with different concentrations (0, 1 mM and 10 mM) for 480 min.The rest of the experimental conditions are the same as described before.
To examine the reusability of the proposed adsorbent, six adsorption-desorption cycles were performed in this work.Adsorption experiments were first conducted by placing the adsorbent (0.1 g) in Cr(VI) solutions (50 mL of 10 mg/L and a solution pH value of 3).The adsorbents adsorbed with Cr(VI) were desorbed using 0.1 M NaOH by shaking.This adsorbent was finally collected by filtration, washed with distilled water, and then reused in the next cycle of adsorption experiments.

XRD analysis
Figure 1a shows the XRD patterns of the CHAP, 0.25CMC-CHAP, 0.5CMC-CHAP, and 1.0CMC-CHAP.As for CHAP, it had typical diffraction peaks of carbonate incorporated hydroxyapatite structure (JCPDS file: #19-0272), [29] 211), ( 112), (300), (310), ( 222), ( 213) and (321) (hkl) reflection planes, respectively. [30,31]In addition, no other secondary phases besides the CHAP structure were detected which suggested that CHAP was successfully synthesized. [32]Compared to pristine CHAP, the CMC-CHAP samples maintained the basic structure of CHAP.With the increasing of CMC content, the full width at half maximum of the (211) peak became wider, manifesting that the CMC-CHAP samples had smaller size and lower crystallinity than CHAP. [33]These results can be ascribed to the inhibitory effect of CMC on the growth of CHAP crystals during the nucleation/crystallization of CMC-CHAP. [27]

FTIR analysis
The FTIR technique was adopted to measure the functional groups on the surface of CHAP and CMC-CHAP samples.The infrared spectrum of CHAP (Figure 1b,i) indicated the carbonate ions peak at about 1456, 1419, 1383 and 874 cm À 1 , [34] which were ascribed to the B-type substitution (carbonate replacing phosphate ions) in HAP crystal lattice, revealing that carbonate groups were incorporated into the apatite structure by substituting of phosphate groups during preparation. [35]Specifically, two broad bands centered in the finger print region 565-610 cm À 1 , and 900-1100 cm À 1 associated to the phosphate group P-O bond.A broad band around 3000-3700 cm À 1 band position was observed due to the O-H groups stretching mode. [36]In comparison, the characteristic bands of CHAP were also shown in all the CMC-CHAP samples (Figure 1b,ii-iv).The relative intensities of the phosphate bands at 565-610 cm À 1 and 900-1100 cm À 1 in 0.25CMC-CHAP, 0.5CMC-CHAP, and 1.0CMC-CHAP samples were slightly decreased with increasing the CMC concentration, indicating that more CMC (Figure 1b,v) was immobilized on CHAP.It was noteworthy that bands at 3567 and 630 cm À 1 were hardly visible after CMC functionalization which indicated that these specific groups on apatite surface were involved in the bridging process with CMC and the low crystallinity of the samples. [37]T surface area analysis The N 2 adsorption-desorption isotherms and the pore size distributions are shown in Figure 1c and 1d, and the corresponding textural properties were also shown in Table 1.The CHAP and CMC-CHAP had similar curve shapes.According to the IUPAC classification, they all belonged to type IV isotherms, which was a typical characteristic of mesoporous material. [38]owever, the textural properties (such as the BET surface area, average pore diameter, and pore volumes) of the samples were varied considerably, depending on the CMC content.As shown in Table 1, the 0.25CMC-CHAP sample had a highest surface area (122.90 m 2 /g) and pore volume (0.545 cm 3 /g) among all the samples.In contrast, the specific surface area and pore volume of bare CHAP were only 56.36 m 2 /g and 0.203 cm 3 /g, as result of the agglomeration of CHAP nanoparticles.Hence the presence of 0.25% CMC could increase the surface area and improve the dispersibility through the bridging process. [27]Nevertheless, with further increasing CMC concentration, the specific surface area and pore volume of 0.5CMC-CHAP and 1.0CMC-CHAP were significantly decreased to be 12.96 m 2 /g (0.0699 cm 3 /g of pore volume) and 7.92 m 2 /g (0.0258 cm 3 /g of pore volume), respectively.This was probably due to the excessive amount of CMC forming a denser CMC coating on the nanoparticles and lowering surface area and pore volumes. [39]Based on the above results, the optimum CMC content was determined to be 0.25%, and the 0.25CMC-CHAP sample would be a potential material to adsorb Cr(VI) contaminants

Morphology analysis
The TEM (Figure 2a-2d) and SEM (Figure 2e-2h) micrographs of the obtained CHAP and CMC-CHAP samples revealed the bare CHAP sample exhibited agglomerated rod-like morphology with nanometer size.In case of the 0.25CMC-CHAP sample (Figure 2b and 2f), the presence of CMC improved the dispersibility of CHAP.It was visible that CMC could suppress the growth and aggregation of CHAP nanoparticles (Figure 2b).Meanwhile, the EDS spectra (Figure 2i-2l) showed that the O content increased with the increasing of CMC concentration in CMC-CHAP samples, confirming the attachment of CMC on CHAP surface.Numerous work reported similar results, which demonstrated that the coexisting CMC and other biomaterials could prevent the agglomeration of chlorapatite, [27] FeS, [40] and zero valent iron [41] nanoparticles.It was considered that negatively charged CMC could attach to CHAP nanoparticles surface through electrostatic interaction and the bidentate bridging of COOH groups in CMC molecule with OH groups of CHAP. [42]owever, as the CMC concentration increased from 0.025% to 1%, the surface of CHAP particles was covered with a denser layer of the CMC molecules (Figure 2c and 2d), leading to complete stabilization of the CHAP particles. [27,28]Since the fully stabilized nanoparticles were not easy to be separate due to their highly stable features, [43] the 0.25% CMC bridged CHAP sample (noted as 0.25CMC-CHAP) was selected for the subsequent experiments.

Effect of pH
The initial pH of the solution in water treatment is an essential factor during the batch mode adsorption process that modifies the surface charge and characteristic property of adsorbent and adsorbate species. [44]The adsorption efficiency of Cr(VI) onto 0.25CMC-CHAP sample as a function of pH (2-10) was studied and results are presented in Figure 3a.It was obvious that, the adsorption of Cr(VI) was highly pHdependent, and the maximum removal rate for Cr(VI) appeared at pH 2-3, and then with the increase of the pH value, the adsorption of Cr(VI) gradually decreased.By increasing the pH values from 2 to 6, the adsorption of Cr(VI) on CMC-CHAP nanoparticles decreased from 4.55 mg/g (91.1% adsorption efficiency) to 3.27 mg/g (65.4% adsorption efficiency), respectively.Generally, Cr(VI) is negatively charged in aqueous solution under all pH conditions. [45]The dominant forms at which adsorption occurs are HCrO 4 -and Cr 2 O 7 2-, which formed between pH 2 and 6. [46] Therefore, the high Cr(VI) adsorption at low pH is attributed to the strong electrostatic attraction between the Cr(VI) oxyanions and the positive charged surface of adsorption material. [11]With the pH increased, the electrostatic repulsion between Cr(VI) and negatively charged surface of adsorbent became strong, resulting in the declined adsorption efficiency. [47]Furthermore, the significant decrease in Cr(VI) adsorption after pH > 7 could be attributed to the competition between Cr(VI) anions and hydroxyl groups, which blocked the large numbers of adsorbent active sites and significantly reduced the adsorption capacity. [24,25]Similar phenomenon has been reported for Cr(VI) adsorption using apatite based materials, such as HAP, [22] CHAP, [25] and fluorapatite. [48]

Effect of adsorbent dosage
The effect of adsorbent dose on Cr(VI) adsorption by CMC-CHAP is presented in Figure 3b.It could be seen from Figure 3b that Cr(VI) adsorption efficiency increased significantly with increase in adsorbent dose, when the adsorbent dose was 0.1 g/50 mL, nearly 85% of Cr(VI) was removed.The increase in the efficiency of adsorption was attributed to the fact that more surface area and adsorption sites were available for Cr(VI) to be adsorbed with increase in adsorbent dose. [49]However, the unit adsorption density was decreased from 7.06 to 0.32 mg/g as the dose was increased from 0.025 to 0.25 g/50 mL.This could be owing to the unsaturation of adsorption sites during adsorption process. [50]Further increase of adsorbent dosage over 0.15 g/ 50 mL, the residue Cr(VI) concentration was less than 0.5 mg/L after adsorption, in line with the Chinese National Sewage Comprehensive Emission Standard (GB8978-1996). [51]The final removal rates reached 93.8%, 96.1% and 96.8% when the dosage was 0.15, 0.2 and 0.25 g/50 mL, respectively.These results proved that CMC-CHAP is a high-efficient adsorbent for Cr(VI) removal.

Adsorption kinetics
Adsorption kinetics was an important constant for the evaluation of a good adsorbent.As shown in Figure 3c, the Cr(VI) uptake on CMC-CHAP was rapid in the first 30 min, contributing to about 65.8%, 69.2% and 68.8% of the ultimate adsorption amount for 5, 10 and 20 mg/L Cr(VI), respectively, and then augmented gradually.In present study, the adsorption equilibrium was achieved within about 6 h.Analogously, Yu et al. [52] demonstrated a rapid uptake of Cr(VI) ions by sepiolite-supported magnetite nanoparticles at the first 2 h and that the equilibrium was attained within 12 h.This phenomenon occurred because plenty of empty surface sites were available for Cr(VI) capture during the early stage and after some time the remaining adsorption sites were hard to be occupied due to the repulsive force between the Cr(VI) species on the solid surface and in bulk phase. [53]e analysis of adsorption kinetics is a significant tool for analyzing the adsorption process, it provides valuable experimental information for designing water treatment system. [24]The kinetic experimental data of Cr(VI) adsorption onto CMC-CHAP were fitted to the kinetic models: PFO and PSO and intraparticle diffusion models.The kinetic fitting plots were shown in Figure 3c and 3d, and the constants of the three models are listed in Tables S1 and S2.It was observed that the correlation coefficients (R 2 ) for PSO kinetic model (>0.95) were higher and much closer to unity compared to those found with the PFO model (0.82 < R 2 < 0.93).Additionally, the q e values calculated from the PSO model were in good agreement with the experimental values.These finding indicated that the PSO model described well the adsorption of Cr(VI) onto CMC-CHAP.This model suggested that the chemisorption involving sharing or electron exchange between metal ions with adsorbent was the rate limiting step. [54]he intraparticle diffusion model was further applied for judging the rate-controlling step during the adsorption process of Cr(VI) on the CMC-CHAP, as depicted in Figure 3d and Table S2.Results showed that the adsorption of Cr(VI) (5, 10, and 20 mg/L) could be divided into three stages.The multi-linearity of the plot suggested that the adsorption process of Cr(VI) took place in multi-steps. [55]The first sharper portion was attributed to the diffusion of Cr(VI) through the solution to the external surface of the adsorbent (external diffusion) and the second portion was intra-particle diffusion as the rate-limiting step, in which Cr(VI) gradually entered into the internal pores of CMC-CHAP; while the third stage established equilibrium when the active sites were almost occupied. [56]The multilinearity of the q e versus t 0.5 plots indicated that both external mass transfer and intraparticle diffusion have been involved in different phases of the adsorption process.Figure 3d also demonstrated that line segments fitted with different initial Cr(VI) concentrations all exhibited good linear relationship, but none of them passed through the origin, suggesting that the intraparticle diffusion was not the sole rate-controlling mechanism. [57]From Table S2, it was seen that the order of adsorption rate followed the order: k p1 > k p2 > k p3 , suggesting the fast adsorption at the initial stage due to the external diffusion.Meanwhile, the rate constant values increased significantly with increasing initial Cr(VI) concentration.This was due to the fact that the increase of adsorbate concentration resulted in increase of the driving force, which could enhance the diffusion rate of Cr(VI) species in pore. [58]

Adsorption isotherms and thermodynamic analysis
To evaluate the interaction between an adsorbate and an adsorbent and to design and operate an adsorption system successfully, equilibrium adsorption isotherm data is very important. [59]For the determination of the adsorption thermodynamics of Cr(VI), the isotherms of Cr(VI) adsorption on CMC-CHAP were obtained at different temperatures (283, 298 and 313 K).The non-linear fitting plots and the calculated parameters are presented in Figure S1a-c and Table S3.As shown in Figure S1a, the temperature rise led to an increase in Cr(VI) adsorption capacity.The same dependence for Cr(VI) adsorption by sorbent modified by nano-HAP embedded gelatin [19] and Fe 2 O 3 -Ag nanocomposite. [60]From Table S3 based on the R 2 values, it was concluded that Cr(VI) adsorption on CMC-CHAP better followed Langmuir and Sips models than the Freundlich model, which indicated that the adsorption of Cr(VI) on the surface of adsorbent was composed of both homogeneous and heterogeneous adsorption. [24]The applicability of Langmuir model suggested that there was effectively monolayer adsorption of Cr(VI) on the surface of CMC-CHAP.The parameter ) is a dimensionless parameter which represents the important characteristics of Langmuir model.The values of R L for Cr(VI) adsorption on CMC-CHAP ranged between zero and one (0 < R L < 1) consistent with the requirement for a favorable adsorption process. [57]In the Freundlich model, the fitting results also revealed a thermodynamically favorable adsorption of Cr(VI) on the CMC-CHAP as the Freundlich constants 1/n were all less than 1.Furthermore, the heterogeneity factor (1/n) in the Sips model decreased from 2.4093 to 0.5557 as the temperature increased from 283 K to 313 K, suggesting the decreased homogeneity at higher temperatures.The similar result was reported for the adsorption of Cr(VI) on graphene oxide-Fe 3 O 4 composite. [61]Based on the fitting results of Langmuir model, the maximum adsorption capacity (q m ) of Cr(VI) on CMC-CHAP at 298 K was 13.45 mg/g, which was greater than most of the apatite based adsorbents described in literature, as depicted in Table 2.
The experimental data obtained at different temperatures were used to calculate the thermodynamic parameters of Cr(VI) on CMC-CHAP.To calculate the values of the parameters (DG 0 , DH 0 , and DS 0 ) the following equations were used: where R is the universal gas constant (8.314J/(mol�K)); T is the temperature in Kelvin; K 0 is the thermodynamic equilibrium constant, which was determined by the intercept of the straight plotting (Figure S1d, ln q e /C e versus q e ). [62]he obtained thermodynamic parameters at various temperatures are shown in Table S4.The negative values of DG 0 for all experimental temperatures showed that the adsorption of Cr(VI) was spontaneous and feasible.And the DG 0 values reduced with rising temperature, indicating an enhanced spontaneity with temperature.The value of DH 0 was positive which suggested that the Cr(VI) adsorption process was endothermic in nature and the high temperature promoted Cr(VI) adsorption, which was consistent with the result in isothermal discussion. [17]The positive value of DS 0 exhibited an increment in the disorder at the Cr(VI) ions solution/adsorbent interface, suggesting a good affinity between Cr(VI) and CMC-CHAP surface. [63]fects of coexisting ions There are many common ions exist in wastewater and can compete with Cr(VI) for the adsorption sites.In this work, Cl -, NO 3 -, SO 4 2-, K þ , Ca 2þ , Mg 2þ were selected as interfering anions and cations to explore the influence of interference ions and ionic strength on Cr(VI) removal by CMC-CHAP.As depicted in Figure 4, different concentrations of coexisting ions possessed various degrees of inhibitive influence on adsorption efficiency (%) of Cr(VI).It was found that the coexisted K þ , Ca 2þ , or Mg 2þ had no significant influence on Cr(VI) adsorption.The possible explanation might be that, the positively charged adsorbent surface led to the repulsion with the competing cations. [64]In the case of Cl -, NO 3 -, and SO 4 2-, the competitive adsorption between coexisting anions and Cr(VI) could make the available adsorption sites insufficient to uptake more Cr(VI), thus hindered the adsorption of Cr(VI) to some extent. [65]Even so, approximately > 77% adsorption efficiency was still remained by CMC-CHAP, suggesting that the proposed adsorbent possessed favorable practicability in complex water environment.

Reusability of adsorbent
Reuse of the adsorbents is important due to economic and resource considerations.The selection of an efficient eluent for desorption process is an essential aspect in regeneration study.As acidic pH (pH 2-3) was favorable for adsorption of Cr(VI), desorption of Cr(VI) from saturated CMC-CHAP was performed using 0.1 M NaOH solution as a basic desorbent via shaking. [17,63]Generally, desorption of Cr(VI) at basic condition can take place due to exchanging of CrO 4 2-(the dominant form of Cr(VI) in alkaline solution) with OH -, [66] and the electrostatic repulsion between negatively charged sites and Cr(VI) also facilitate the desorption. [67]Six sequential adsorption-desorption cycles were conducted to investigate the reusability of CMC-CHAP for Cr(VI) removal.It was noticed from Figure S2 that the adsorption efficiency of CMC-CHAP to remove Cr(VI) from aqueous solutions was still kept about 76.5% after six cycle uses.Therefore, it was concluded that CMC-CHAP has excellent reusability, and subsequently represents an alternative adsorbent for effective clean-up of wastewater containing Cr(VI) ions.

Characterization of Cr(VI)-loaded adsorbent
In order to understand the Cr(VI) adsorption mechanism by CMC-CHAP, the Cr(VI)-loaded adsorbent was further characterized using x-ray photoelectron spectroscopy (XPS).As illustrated in Figure 5a, the survey XPS spectrum comprised peaks due to C 1 s, O 1 s, Ca 2p, P 2p and Cr 2p, confirming Cr(VI) adsorption on the CMC-CHAP surface.
The Cr 2p high resolution spectrum (Figure 5b) indicated the presence of two peaks, corresponding to Cr 2p 1/2 and Cr 2p 3/2 , respectively.Based on the peak area analysis, most of the adsorbed Cr(VI) was conversed to Cr(III) (75.6%) with only 24.4% remaining as Cr(VI).The optimum fitting was achieved by deconvoluting the Cr 2p spectrum into four peaks.The peaks at 590.0 and 588.1 eV were corresponded to the Cr 2p1/2 of Cr(VI) and Cr(III), while those around  579.1 and 581.3 eV were assigned to the Cr 2p3/2 of Cr(VI) and Cr(III), showing that and Cr(III) species coexisted on adsorbent surface. [68]These results were in agreement with previous literature.Thus, the prepared CMC-CHAP possessed the ability for the adsorption and partial reduction of Cr(VI) and then adsorption of the generated Cr(III).The reduction reaction of Cr(VI) ion to Cr(III), occurring on adsorbent surface in acidic environment, could be expressed as follows: : However, after the adsorption assays, it was found that Cr(III) ions are not detectable in the solution, which was due to the surface complexation on CMC-CHAP surface. [27,69]ccording to the discussion above and the results of batch experiments, it could be inferred that Cr(VI) was removed by CMC-CHAP mainly through three stages (Figure 6): (1) the predominant Cr(VI) species (HCrO 4 -) was adsorbed by the protonated groups of CMC-CHAP through electrostatic attraction; (2) the adsorbed Cr(VI) were partially reduced to Cr(III) by oxygen containing groups; (3) Cr(III) was further adsorbed on the surface through the strong coordination interactions with the electron rich CMC-CHAP surface functionalities.

Conclusions
In this work, CMC-CHAP was successfully synthesized, characterized and used as a novel adsorbent for the removal of Cr(VI) from aqueous media.It was showed that the combination of CHAP nanoparticles with CMC made it convenient to use practically for Cr(VI) adsorption and detoxification.Results from batch adsorption measurements revealed that the synthesized CMC-CHAP had a promising potential for competing with conventional absorbents, as it exhibited higher adsorption capacity (13.45 mg/g) compared to previously reported adsorbents such as HAP (1.382-2.72 mg/g), [39][40][41] HAP-chitin/chitosan composite (2.845 and 3.45 mg/g), [40] diatomite-modified MnO 2 (3.26 mg/g), [62] activated fluorapatite (0.87 mg/g), [67] and fish scales derived HAP (8.156 mg/g). [73]According to kinetic adsorption analyses, the PSO model best fitted the adsorption that happened and Langmuir isotherm model was favorable with higher correlation coefficient (>0.99).In particular, the coexisting ions had neglectable impact on Cr(VI) elimination, showing the strong selective adsorption capability and anti-interference ability of CMC-CHAP adsorbent.Adsorption-desorption experiments indicated that the CMC-CHAP exhibited satisfactory reusability.The mechanisms of Cr(VI) removal by CMC-CHAP mainly included electrostatic interactions, chemical reduction and surface chelation.These results showed that CMC-CHAP had good application prospects for Cr(VI) removal in environmental pollution cleanup, and CMC-bridging could offer a new angle to extend the use of nanoparticles for water treatment.

Figure 5 .
Figure 5.The XPS total survey spectra (a) and the of the Cr 2p spectra (b) of Cr(VI)-loaded CMC-CHAP.

Table 1 .
BET surface area and pore characteristics of CHAP and CMC-CHAP samples.

Table 2 .
The comparison of maximum Cr(VI) adsorption capacities of apatite adsorbents based on the Langmuir isotherm model.