Removal of Copper (II) from Aqueous Solution by a Hierarchical Porous Hydroxylapatite-Biochar Composite Prepared with Sugarcane Top Internode Biotemplate

Abstract

Porous hydroxyapatite-biochar composites with layered microstructures (SC–HA/C) were prepared by carbonizing sugarcane stem nodes and then soaking them in lime water and (NH₄)₂HPO₄ solutions in rotation. The surface area of SC–HA/C ranges from 8.52 to 28.44 m²/g, and its microstructure inherits various macro-, meso-, and micro-pores in the cell walls of sugarcane and in the pits of the vessel walls. The maximum removal capacities were 11.50, 14.65, and 19.81 mg/g for the Cu (II) immobilization at 25 °C, 35 °C, and 45 °C with the solution Cu (II) concentration of 10~320 mg/L, respectively, which were in accordance with the copper sorption capacities of synthesized nano-hydroxylapatites. The Cu (II)-removal kinetics and isotherm followed the pseudo-second-order equation and the Langmuir equation very well. The formation of the Cu-containing hydroxylapatite solid solutions ((CuxCa_1−x)₅(PO₄)₃(OH)) through adsorption, ion exchange (x = 0.01~0.04), and dissolution-coprecipitation (x = 0.13~0.35) was the dominant process for the Cu (II) removal by the SC–HA/C composite.

1. Introduction

Copper is not only a drinking-water pollutant but also an essential element. Many mining and metal industries, such as metal cleaning and electroplating, battery manufacturing, petroleum, wood pulp production, dye, and the fertilizer industry, release copper at undesired levels [1,2,3]. Copper is frequently detected in surface water and wastewaters, resulting in a potential and significant threat to biodiversity of the ecosystem and human health [4,5,6]. Copper has been reported to cause liver damage, lung cancer, jaundice, neurotoxicity, vomiting, and diarrhea [6,7,8,9]. A copper excess can cause plant growth retardation [6].

Water containing Cu > 1.3 mg/L can lead to lung cancer and liver damage of human beings [8,10]. The China Standards for Drinking Water Quality recommended the level of 1 mg/L Cu in drinking water. The acceptable limit of 2.0 mg/L Cu in drinking water has been also prescribed by the World Health Organization. Therefore, it is necessary to efficiently reduce the Cu concentrations in various industrial wastewaters to the acceptable limit before their discharge into the environment [4,11]. Many methods, including chemical precipitation, coagulation, adsorption and biosorption, ion-exchange, nanofiltration, membrane processes, and reverse osmosis, have been applied in eliminating copper ions from solutions [3,6]. Among all the available methods, adsorption is considered as one of the most economic and effective techniques [3,8].

Various materials have been applied for the Cu (II) elimination, for instance, sugarcane bagasse [2,3,5,6,12,13], bagasse fly ash [2], modified sugarcane bagasse [3,5], loofa sponge [14], wood sawdust [4], banana peel and watermelon rind [6], bagasse-activated carbon [15], natural phosphate rock [16,17,18], synthesized hydroxylapatite [7,19,20], nano-hydroxylapatite/chitin composite and nano-hydroxylapatite/chitosan composite [1], and carbon nanotube sheets [21].

Hydroxylapatite (HA, Ca₅(PO₄)₃(OH)) can be used to efficiently eliminate metallic cations from polluted waters through adsorption or co-precipitation because of its low cost, large surface area and adsorption capacity, high stability under different redox conditions, and low solubility and leachability of adsorbed pollutants from exhausted adsorbents [11,20,22,23]. However, hydroxylapatite nanoparticles tend to aggregate significantly in the presence of heavy metal cations and are hence difficult to be separated from treated effluents [24]. The pulverized hydroxylapatite cannot be directly applied in flow-through systems, such as fixed-bed column, because it can result in possible clogging, significant hydraulic obstruction, and hence excessive pressure drops [1]. These technological bottle-necks can be solved by coating calcium hydroxylapatite onto hierarchical porous microstructures of plants [25,26], which is helpful for porosity [27].

Sugarcane is widely planted and the most productive crop in the world for both sugar and alcohol industries in Asia and South America [28,29]. It generates a great deal of agricultural residues, including sugarcane tops and bagasse, which is a threat to the environment, and an appropriate disposal of such residual wastes is a primary target to be pursued [28,30]. Sugarcane tops containing less sucrose are habitually discarded in the field after harvesting [29,30,31,32]. The fibrous vascular system of sugarcane is an open network with large surface area and porosity [33]. Sugarcane biomass (bagasse/sugarcane stalk) has great potential for the preparation of hierarchical porous carbon adsorbent materials [34]. The use of sugarcane biomass to prepare porous biomass carbon and its derivatives as adsorbents to remove pollutants from water solves both the solid-waste-disposal problem and the water-pollution problem [26,35].

In this work, sugarcane top residuals were sampled and cut, and then, their stalk internodes were selected as microstructural biotemplate to prepare the hierarchical porous hydroxylapatite-biochar composite (SC–HA/C). The Cu (II)-elimination test was carried out using the prepared SC–HA/C, which took account of the investigation on the influences of removal condition, the removal kinetics, isotherm, and mechanism.

2. Materials and Methods

2.1. Material Preparation and Characterization

2.1.1. Preparation

The preparation method of the material is described in previous research [26,36]. Residual sugarcane tops were first sampled and cut into 1 × 1 × 1 cm³ pieces, from which stalk internodes were then selected and dipped in 5% ammonia solution and boiled for (100 °C) six hours to increase the porosity and pore connectivity by swelling of cellulose, disrupting of lignin and hemicellulose, hydrolyzing of uronic acid esters, and removing of extractable constituents from the sugarcane microstructure [37]. The treated sugarcane tops were thoroughly rinsed in ultrapure water, heated at 80 °C for 1 day, and then heated in a muffle furnace at 350 °C for 3 h (air atmosphere), then cooled to room temperature. Subsequently, the preprocessed sugarcane stalk cubes were soaked with saturated limewater for 2 h and with 0.02 mo1/L diammonium phosphate solution for another two hours in turn for five times to precipitate hydroxylapatite (HA) onto the microstructure of sugarcane biotemplates. Lastly, the soaked biotemplates were oven-dried at 50 °C for twenty-four hours and then air-cooled to obtain the SC–HA/C composite.

2.1.2. Characterization

Phases of the SC–HA/C composite before and after Cu (II) immobilization were determined by recording its powder X-ray diffraction pattern (XRD) with a diffractometer under Cu Kα radiation (X’Pert PRO, PANalytical B.V.). The surface morphology and area of SC–HA/C were observed by a scanning electron microscope (SEM, JEOL JSM-6380LV, Japan Electron Optics Ltd.) with an energy dispersive X-ray spectroscope (EDS, IE350, Oxford Instruments) and a Brunauer–Emmett–Teller (BET) surface area analyzer (NOVAe1000, Quantachrome), respectively. The Thermo Scientific ESCALAB 250Xi apparatus was used to obtain the X-ray photoelectron spectra (XPS) of SC–HA/C before and after Cu (II) removal.

2.2. Batch Experiments

Batch removal tests were made in 100 mL centrifuge tubes for copper removal from synthetic wastewater at a set copper concentration. The influences of reaction time (0~60 h), solution Cu (II) concentrations (5~350 mg/L), temperatures (25~45 °C), initial solution pH (1~8), SC–HA/C doses (0.10~1.00 g/50 mL), and SC–HA/C particles sizes (<100 mesh ~ >20 mesh) were investigated. All solutions after Cu (II) removal were filtered using 0.22 µm micropore membranes, and the remaining copper concentrations were analyzed with an inductively coupled plasma-optical emission spectrometer (ICP-OES, Perkin-Elmer Optima 7000 DV). The amounts of Cu (II) removed q_t (mg/g) at time t were computed according to q_t = (C_o − C_t) V/W, where C_t is the copper concentration at time t (mg/L), V the solution volume (mL), and W the SC–HA/C dose (g). The copper-removal percentages were computed according to Rate (%) = (C_o − C_e)/C_o × 100, where C_o is the initial copper concentrations (mg/L) and C_e the remaining copper concentration (mg/L).

3. Results and Discussion

3.1. Characterization of the Composite

3.1.1. XRD

Both the relative intensity and the peak position of the XRD patterns recorded for SC–HA/C were consistent with hydroxylapatite (Figure 1 and Figure 2). The six strongest reflections, (211), (002), (213), (222), (004), and (210), of hydroxylapatite (Reference code 00-001-1008) corresponded to the six highest diffraction peaks at 2θ = 32.04°, 25.88°, 49.52°, 46.64°, 53.19°, and 28.68° observed for SC–HA/C, respectively. A noticeable structural change after Cu (II) removal for the initial Cu (II) concentration of 0~300 mg/L indicated the formation of the copper-containing hydroxylapatite solid solutions ((Cu_xCa_1−x)₅(PO₄)₃(OH)) through the isomorphic substitution of Cu (II) (0.72 Å) for Ca (II) (0.99 Å) in hydroxylapatite. The slight regular shift of the (211) and (002) peaks to the higher diffraction angles also showed that the Cu/(Cu + Ca) molar ratio (x) increased with the increasing copper concentration up to 300 mg/L (Figure 2). Similar to the properties of cadmium–calcium hydroxylapatite solid solution, the exchange of smaller Cu (II) with larger Ca (II) in the calcium hydroxylapatite lattice leads to structural contraction [38].

3.1.2. SEM and EDS

The characteristic SEM images of the SC–HA/C composite materials in transverse and longitudinal direction are presented in Figure 3A,B, respectively. The SC–HA/C composite retained the honeycomb-like microstructure of sugarcane top stalks, which included micro-, meso-, and macro-pores with a porosity of 94.58% and a BET surface area of 8.52~28.44 m²/g, from the pits of cell and vessel walls, vascular bundles within sclerenchyma sheaths, and vacuolated storage parenchyma cells (Figure 3A). The crushed SC–HA/C composite showed smooth surfaces and no gaps between hydroxylapatite and sugarcane stalk biochar. The interconnected porous structure of SC–HA/C is helpful for the free migration of electrolyte solution [34].

The EDS analysis indicated that the SC–HA/C composite contained Ca, P, O, and C, and the surface of the SC–HA/C composite material after Cu (II) removal had a low Cu/(Cu + Ca) molar ratio of about 0.01~0.35 (Figure 4). Some fine, needle-like precipitates on the SC–HA/C surface after Cu (II) removal were also observed in Figure 4, G. Besides, the porous structure of SC–HA/C after Cu (II) removal showed no visible variation and no clogging.

3.1.3. XPS

XPS analysis was carried out the SC–HA/C composites before and after Cu (II) removal to examine their surface bonding states and compositions. The result indicated that SC–HA/C consisted of carbon mostly from sugarcane top stalk biotemplates, calcium, phosphorus, and oxygen from the soaking solutions (Figure 5).

After Cu (II) removal, the Cu2p peak in the spectra was observed. The shifts of the carbon, calcium, phosphorus, and oxygen peaks after Cu (II) removal were <0.05eV, and therefore, the Cu (II) removal did not affect the phosphorus and calcium states on the SC–HA/C surfaces obviously. From the peak areas of Ca2p, P2p, O1s, C1s, and Cu2p after Cu (II) removal by SC–HA/C, its composition was estimated to be 4.85%, 10.07%, 16.57%, 67.52%, and 0.99% for Ca, P, O, C, and Cu, respectively. The SC–HA/C composite had a Cu/(Cu + Ca) atomic ratio of 0.17 that was in the EDS range of 0.01~0.35.

3.2. Influence of Removal Condition

3.2.1. Influence of Interaction Time

Influence of interaction time on the copper removal by SC–HA/C was studied for initial copper concentrations of 10~50 mg/L at 25 °C (Figure 6). Two different removal stages were detected. The first stage was a rapid Cu (II) elimination, while the second stage was a quantitatively unimportant and slow process. The amount of Cu (II) removed by SC–HA/C increased and the remaining solution Cu (II) concentration declined rapidly with time at the early stage, which indicated a strong interaction between SC–HA/C and copper ions (Figure 6).

The amount of copper removed by SC–HA/C and the remaining Cu (II) concentration were almost stable after 5 min interaction at 25 °C and the initial pH 5.0. For the initial copper concentrations of 10 and 20 mg/L, >98% of them was eliminated in 5 min at the average removal rates of 0.2072 and 0.4011 mg/(g·min), respectively. For the initial copper concentration of 50 mg/L, >95% of them was removed in 45 min at the average rate of 0.1102 mg/(g·min).

3.2.2. Influence of Temperature and Initial Concentration

The amount of Cu (II) removed by SC–HA/C (q_e) vs. the Cu (II) concentration is plotted in Figure 7. The results of the replicate experiments at different initial Cu (II) concentrations are presented in the end of main text (Figure A1).

The copper immobilization at different temperatures showed a similar evolution trend. For the immobilization at 25 °C, the amount of copper removed increased from 0.47 to 11.50 mg/g as the initial copper concentrations were increased from 5 to 250 mg/L; for the removal at 35 °C, the amount of copper removed increased from 0.49 to 14.65 mg/g as the initial copper concentrations were increased from 5 to 210 mg/L; for the removal at 45 °C, the amount of copper removed increased from 0.50 to 19.81 mg/g as the initial copper concentrations increased from 5 to 320 mg/L. However, the amount of copper removed declined with the further increase of the initial copper concentrations. Commonly, the removal rates declined continuously from 93.47% to 23.10% at 25 °C, from 99.25% to 26.70% at 35 °C, and from 99.98% to 53.43% at 45 °C with the increasing copper concentrations. The copper-removal percentage declined with the increasing copper concentrations because of the lack of active sites that were available for the large copper concentrations. The removal capacity of SC–HA/C increased with the increasing copper concentrations because of the increasing mass transfer driving force for copper ions from solution to the SC–HA/C surface. The amount of copper removed (q_e) and the removal rate increased as the temperature was increased, which suggested that the temperature increasing was beneficial for the Cu (II) ions to diffuse more quickly through SC–HA/C, and the Cu (II)-removal process was endothermic [24].

3.2.3. Influence of Initial pH

The copper immobilization by hydroxylapatite is significantly affected by solution pH, which was investigated by agitating 0.5 g SC–HA/C in 50 mL 10~50 mg/L Cu (II) solution over a pH range of 1~6 at 25 °C. The amount of Cu (II) removed by SC–HA/C (q_e) increased with a rising pH up to 3~5 and attained the steady state at pH 4~6, indicating the presence of two removal processes (Figure 8).

Both the Cu (II) solution chemistry and the SC–HA/C surface charge might be greatly influenced by system pH. The attaching of positively charged copper ions onto SC–HA/C was largely influenced by the SC–HA/C surface charges that were dependent on the solution pH. Higher solution pH benefited copper removal by SC–HA/C. For the copper concentration of 10~50 mg/L, most of copper cations were eliminated from solution and the removal rates achieved 98.22~99.85% at pH 5~6.

3.2.4. Influence of SC–HA/C Dose

Influence of the SC–HA/C dose (0.10~1.00 g/50 mL) on the Cu (II) removal from solution at 25 °C and initial pH 5.0 for 24 h was investigated. The amount of copper removed by SC–HA/C increased with the increasing dose (Figure 9), which was relevant to the increasing surface area and so the increasing accessibility of active sites.

With the increasing SC–HA/C dose from 0.10 to 1.00g in 50 mL solution, the amount of copper removed declined steadily from 5.79 to 0.58 mg/g, 11.41 to 1.16 mg/g, and 20.30 to 2.66 mg/g; the Cu (II)-elimination efficiencies increased continuously from 99.75% to 99.95%, 98.69% to 99.87%, and 76.39% to 99.74% for the copper concentrations of 10, 20, and 50 mg/L, respectively.

3.2.5. Influence of SC–HA/C Particle Size

The Cu (II)-removal capacity and percentage were dependent on the SC–HA/C particle size at given copper concentrations. The amounts of copper removed by the uncrushed SC–HA/C (>3 mm) were 1.10, 2.13, 5.09, and 8.77 mg/g with the removal rates of 99.43%, 99.37%, 97.47%, and 79.20% for the Cu (II) concentrations of 10, 20, 50, and 100 mg/L, respectively (Figure 10).

The crushed SC–HA/C (<100 mesh) and the uncrushed SC–HA/C showed a larger elimination percentage than the sugarcane top stalk biochar and the biochar–HA mixture at different Cu (II) concentrations. For the copper concentration of 10~100 mg/L, neither the crushed SC–HA/C (<100 mesh) or the un-crushed SC–HA/C had a high removal capacity and a large elimination efficiency that was comparable to HA nano-particles. For the copper concentration of 10~100 mg/L, the smallest removal capacity was 1.10~6.64 mg/g of the crushed SC–HA/C (40~60 mesh) with the removal percentage of 60.02~99.34%. The largest removal capacity was 1.10~10.70 mg/g of the crushed SC–HA/C (<100 mesh) with the removal percentage of 96.62~99.91%. Relatively, the un-crushed SC–HA/C (chip, >3 mm) also showed a large removal capacity (1.10~8.77 mg/g) and a large removal percentage (79.20~99.43%) due to the special porous microstructure of SC–HA/C.

3.3. Removal Kinetics and Isotherm

3.3.1. Kinetics

A kinetic test was performed by agitating 0.5 g of SC–HA/C in 50 mL solution of 10~50 mg/L Cu (II) at 25 °C and pH 5.0. The solutions were sampled at different time intervals from 5 to 1440 min. The experimental result was fitted to different kinetic models, among which the pseudo-second-order kinetic equation presented the best fitting with the correlation coefficient R² = 1.0000 (Figure 11).

The equation is described as

$$\frac{t}{q_t}=\frac{1}{k_2{q}_e^2}+\frac{t}{q_e}$$

(1)

where k₂ is the rate constant of the model (g/(mg·min)), q_t the amount of removal time t (mg/g), and q_e the amount of removal equilibrium (mg/g). The removal rate h is defined as

h (mg/(g min)) = k₂q_e²

(2)

The k₂ values were calculated to be 0.1115~2.1613 g/(mg·min) and the h values to be 2.3491~7.0771 mg/(g·min) (Table S1). The estimated removal capacities (q_e), 1.04, 2.02, and 5.09 mg/g, agreed very well to the measured data of 1.04, 2.02, and 5.09 mg/g for the initial copper concentrations of 10, 20, and 50 mg/L, respectively.

3.3.2. Isotherm

The Cu (II)-removal isotherm was studied under the condition of initial copper concentration of 5~350 mg/L, 25~45 °C, initial pH of 5, and SC–HA/C dose of 0.5 g/50 mL.

The experimental result was fitted to the Langmuir equation (Figure 11) that is described as

C_e/q_e = 1/(q_m·K_L) + C_e/q_m

(3)

where K_L is the Langmuir constant (L/mg), C_e the equilibrium concentration (mg/L), and q_m the maximum removal of Cu (II) (mg/g). K_L was computed from the linear intercept in the C_e/q_e vs. C_e plot and q_m from the slope (Table S1).

Furthermore, the favorability of Cu (II) removal by SC–HA/C was studied through R_L, a dimensionless separation factor, that can be described as

R_L = 1/(1 + K_LC_o)

(4)

The Langmuir isotherm yielded a very good fit for the copper immobilization by SC–HA/C with the correlation coefficient (R²) of 0.9720 at 25 °C, 0.9713 at 35 °C, and 0.9956 at 45 °C (Figure 11 and Table S1). The large R² showed the homogeneous property of the SC–HA/C surface, where each Cu (II) ion is attached onto the SC–HA/C surface with identical activation energies. The R_L values of 0.0039~0.4044 meant a favorable Cu (II) removal by SC–HA/C. The Langmuir isotherm equation fitted well the Cu (II)-removal data, indicating a possible ion exchange on the hydroxylapatite surface [24] or a monolayer adsorption [19].

3.4. Comparison of Copper-Removal Capacities with Other Materials

The Cu (II)-removal capacities of the SC–HA/C composite together with sugarcane bagasse, bagasse activated carbon, hydroxylapatite (HA), and other materials used for Cu (II) decontamination in the literature are listed in Table S2. The Cu (II)-removal capacity of SC–HA/C was larger than those materials prepared from sugarcane bagasse. The Cu (II)-removal capacities were determined to be 4.13~10.64 mg/g for natural sugarcane bagasse [3,5,6,12,13,39,40,41], 0.19~0.27 mg/g for sugarcane biochar (this study), 1.10~3.81 mg/g for activated carbon (this study), 13.24 mg/g for bagasse-activated carbon [9], 11.87 mg/g for NaOH-modified bagasse [5], 13.34 mg/g for tetraethylenepentamine-modified sugarcane bagasse [3], 5.12~8.24 mg/g for some agricultural residuals [4,6,14], 1.998~16.8 mg/g for natural phosphate rocks [16,17,18], and 10.08 mg/g for carbon nanotube sheets [19].

The synthesized hydroxylapatites (HA) showed a similar Cu (II)-removal capacity under similar test conditions. The largest Cu (II)-removal capacities of some commercial hydroxylapatites were 17.79 mg/g for the synthesized HA from Merck (Fernane et al., 2006), 7.63 mg/g for the synthesized HA from Bio-Rad [19], and 1.02~48.55 mg/g for the synthesized HA from Alfa Aesar [7]. The Cu (II)-removal capacity of SC–HA/C was 0.47~19.81 mg/g, which was larger or comparable to the Cu (II)-removal capacities for the synthesized nano-HA adsorbents, such as 4.7 mg/g for the synthesized nano-HA, 5.4 mg/g for the nano-HA/chitin composite, and 6.2 mg/g for the nano-HA/chitosan composite [1].

3.5. Removal Mechanism

The Cu (II)-removal mechanism by SC–HA/C is a complex process. There is a significant difference in the ionic radius of Cu²⁺(0.72 Å) and Ca²⁺(0.99 Å) and raises the question whether an isomorphous substitution can be obtained when Ca²⁺ is replaced by Cu²⁺ in Ca₅(PO₄)₃(OH). The calcium hydroxylapatite removed heavy metals from solution through adsorption and ion exchange, and no significant change could be observed in XRD analyses [42]. The chemical formula of hydroxylapatite became (Cu_xCa_1−x)₅(PO₄)₃(OH) after ion exchange [42]. The results of this work agreed well with the previous research [7]. The XRD and SEM results showed no secondary crystalline phase other than HA was detected and no differences in the morphology of the solid residues after Cu (II) immobilization by SC–HA/C, while the EDS and XPS analyses indicated the Cu presence on the SC–HA/C surface and the (Cu_xCa_1−x)₅(PO₄)₃(OH) formation (Figure 1, Figure 2, Figure 3, Figure 4 and Figure 5).

The mechanism for the Cu (II) immobilization by HA was greatly depended on experimental conditions [1,7,11,22,43,44]. In the present study, the possible mechanism for Cu immobilization by SC–HA/C included adsorption and formation of surface complexes, ion exchange through Cu²⁺-for-Ca²⁺ substitution in HA, and dissolution/coprecipitation with replacement of Ca²⁺ by Cu²⁺ during recrystallization (Figure 12).

(a)

Adsorption and formation of surface complexes

Copper was effectively immobilized in aqueous solution by SC–HA/C through a two-step process, including the surface complexing of Cu²⁺ onto the POH≡ sites and the Cu²⁺ exchange with Ca²⁺ and causing the Cu-containing hydroxylapatite formation [22]. Cu (II) uptake resulted in a decline of solution pH [7], and H⁺ ions were released from HA surface groups during the Cu (II) adsorption after the following reaction (Figure 12a):

Ca₅(PO₄)₃–OH + Cu²⁺ = Ca₅(PO₄)₃–O–Cu⁺ + H⁺

(5)

The percentage distribution of all hydrolyzed Cu (II) species depended on the solution pH, which was computed using the PHREEQC program with its phreeqc.dat database [45] and plotted together with the experimental data in Figure 13. The Cu (II) removal by SC–HA/C increased with rising pH in acidic solution (Figure 8).

The positively charged Cu (II) species dominated at lower pH and so could be adsorbed onto SC–HA/C quickly. However, several Cu (II) species of various charges, including Cu₂(OH)₂²⁺, Cu (OH)⁺, Cu (OH)₂⁰, Cu (OH)₃⁻, and Cu (OH)₄²⁻, could be present at higher pH.

2Cu²⁺ + 2H₂O = Cu₂(OH)₂²⁺ + 2H⁺   log_k = −10.594

(6)

Cu²⁺ + H₂O = Cu (OH)⁺ + H⁺             log_k = −7.497

(7)

Cu²⁺ + 2H₂O = Cu (OH)₂⁰ + 2H⁺       log_k = −16.194

(8)

Cu²⁺ + 3H₂O = Cu (OH)₃⁻ + 3H⁺      log_k = −26.879

(9)

Cu²⁺ + 4H₂O = Cu (OH)₄^2- + 4H⁺       log_k = −39.98

(10)

The immobilization of Cu (II) occurred on the surface M(2) sites of hydroxylapatite, while the adsorption onto the M(1) sites of the inner structure was not likely. The M(2) sites could principally be more suitable to host copper ions since they exhibited slightly shorter Cu–O bonds in the HA structure channels and were more easily accessed by Cu (II) [22].

• (b) Ion exchange process through substitution of Cu (II) for Ca (II) in HA

The Cu (II) removal could be performed through ion exchange on the SC–HA/C surface, resulting in an increasing solution calcium concentration with time (Figure 6) and the formation of the Cu-containing hydroxylapatite solid solutions ((Cu_xCa_1−x)₅(PO₄)₃(OH)) with a lower Cu/(Cu + Ca) molar ratios (x = 0.01~0.04) (Figure 12b) according to the following general reaction:

Ca₅(PO₄)₃(OH) + 5x Cu²⁺ = 5x Ca²⁺ + (Cu_xCa_1−x)₅(PO₄)₃(OH)

(11)

The solution Ca (II) concentration increased with the decreasing Cu (II) concentration (Figure 8), and the relation between the amount of Cu (II) removed by SC–HA/C and the amount of Ca (II) released was obviously linear with the coefficients (R²) >0.99 (Figure 14), which appeared to be in accordance with the Cu–Ca ion exchange (Figure 12b).

However, it is difficult to determine the exact contribution of the ion-exchange mechanism to the Cu (II) elimination because the HA dissolution could also partly release Ca (II) ions into solution [7,46]. Additionally, the hydration process could affect both the adsorption and the Cu–Ca ion exchange on M(1) and M(2) sites of the HA surface, but hydroxyl channels and surface phosphate groups would not be decomposed. Cu (II) ions were also found to occupy the M(2) vacancy sites more favorably due to less spatial constraint [47].

• (c) Dissolution of HA and precipitation with the Cu (II)-for-Ca (II) replacement during HA recrystallization

The process for Cu (II) removal at low pH also involves the partial dissolution of calcium hydroxylapatite and the following coprecipitation of the Cu-containing hydroxylapatite solid solutions ((Cu_xCa_1−x)₅(PO₄)₃(OH)) (Figure 12c), which was the main mechanism and overwhelmed surface adsorption and ion exchange [46,48,49]. The overall process is pH dependent, and so, the following protonation-deprotonation reactions during the HA dissolution should be considered:

Ca₅(PO₄)₃(OH) = 5Ca²⁺ + 3PO₄³⁻ + OH⁻ log_k = −58.333

(12)

Ca²⁺ + H₂O = Ca (OH)⁺ + H⁺    log_k = −12.697

(13)

Ca²⁺ + 2H₂O = Ca (OH)₂ + 2H⁺    log_k = −22.804

(14)

PO₄³⁻ + H⁺ = HPO₄²⁻    log_k = 12.375

(15)

PO₄³⁻ + 2H⁺ = HPO₄⁻    log_k = 19.573

(16)

PO₄³⁻ + 3H⁺ = HPO₄⁰     log_k = 21.721

(17)

Ca²⁺ + PO₄³⁻ = CaPO₄⁻     log_k = 6.46

(18)

Ca²⁺ + H⁺ + PO₄³⁻ = CaHPO₄     log_k = 15.035

(19)

Ca²⁺ + 2 H⁺ + PO₄³⁻ = CaH₂PO₄⁺    log_k = 20.923

(20)

These reactions resulted in the increase in the solution calcium and phosphate concentrations simultaneously (Figure 6). With the dissolution of calcium hydroxylapatite, the Cu-containing hydroxylapatite solid solutions ((Cu_xCa_1−x)₅(PO₄)₃(OH)) with a higher Cu/(Cu + Ca) molar ration (x = 0.13~0.35) precipitated after the following reaction:

5x Cu²⁺ + (5 − 5x) Ca²⁺ + 3PO₄³⁻ + OH⁻ = (Cu_xCa_1−x)₅(PO₄)₃(OH)

(21)

Nucleation and growth process can be significantly affected by the system saturation state [50]. HA was unlikely dissolved at a high degree after Equation (12) because of the low-solubility product of HA (K_sp = 10^−57.75 for calcium hydroxylapatite). It is more likely that the Cu (II) immobilization was a continuous dissolution/precipitation process [46]. Copper phosphates (for example, Cu₅(PO₄)₃(OH), K_sp = 10^−51.6; Cu₅(PO₄)₃Cl, K_sp = 10^−54.0) also possess much smaller solubility products than other copper compounds [51], showing lower leachability, bioavailability, and toxicity of copper in phosphates. The saturation indices were calculated using the PHREEQC (pH-Redox-Equilibrium Chemistry, version 3) program (Figure 15). The result indicated that the solution of the copper concentration <50 mg/L was only over-saturated with respect to Ca-hydroxylapatite and always under-saturated with respect to other possible calcium and copper compounds, including Cu₅(PO₄)₃(OH).

The HA-dissolution process decomposed hydroxyl channels and surface phosphate groups and affected the following Cu (II) occupation on M(1) and M(2) sites in the formation of the Cu-containing hydroxylapatites ((Cu_xCa_1−x)₅(PO₄)₃(OH)). Energetically, Cu (II) ions preferred to occupy the M(2) vacancy sites and Ca (II) ions the M(1) vacancy sites due to less spatial constraint [22].

4. Conclusions

The prepared porous SC–HA/C composite retained the hierarchical microstructures of sugarcane stalks with the porosity of 94.58% and the surface area of 8.52~28.44 m²/g. The highest removal capacities were determined to be 11.50, 14.65, and 19.81 mg/g for the performance at 25 °C, 35 °C, and 45 °C with the initial copper concentration of 10~320 mg/L, respectively, which were in accordance with the copper-sorption capacities of synthesized nano-hydroxylapatites. The Cu (II)-removal kinetics and isotherm followed the pseudo-second-order equation with the regression coefficients (R²) of 1.0000 and the Langmuir equation with R² > 0.9713. The Cu (II) removal by SC–HA/C was governed by the following reactions, i.e., the Cu ions adsorption onto the SC–HA/C surface, the Cu-for-Ca ion exchange and the formation of the Cu-containing hydroxylapatite solid solutions (Cu_xCa_1−x)₅(PO₄)₃(OH) (x = 0.01~0.04) on the SC–HA/C surface, and the partial dissolution of hydroxylapatite and the following coprecipitation of the Cu-containing hydroxylapatite solid solutions ((Cu_xCa_1−x)₅(PO₄)₃(OH)) (x = 0.13~0.35). Currently, SC–HA/C composites are still in the stage of theoretical exploration. In order to promote the large-scale application of this scheme, on the one hand, it is necessary to further reduce the cost of material fabrication and improve its reproducibility and copper recycling; on the other hand, the removal mechanism of the coexistence of multiple pollutants needs to be further explored.