Sage Journals: Discover world-class research

Abstract

This paper evaluates the pyrolysis conditions applied during the synthesis of bovine bone char and the effect of these parameters in its adsorption properties for heavy metal removal from aqueous solution at batch reactors. The synthesis route has been analyzed in detail and the surface interactions involved in the adsorption process has been studied and discussed using different characterization techniques, given a special emphasis on X-ray photoelectron spectroscopy (XPS) analysis. A proper selection of the pyrolysis conditions improved the metal uptake of bone chars by 143% where the adsorption capacities ranged from 68.3 up to 119.4 mg/g. The removal performance followed the trend Cd²⁺ > Zn²⁺ > Ni²⁺. However, the multicomponent removal of Zn²⁺, Cd²⁺, and Ni²⁺ ions in both binary and ternary mixtures was a strong antagonistic adsorption process. XPS analysis confirmed that the ion exchange between the calcium, from the hydroxyapatite structure of bone char, and the heavy metals in solution played an important role in the adsorption process. These findings are useful to enhance the efficacy of heavy metal removal from aqueous solution using bone char.

Keywords

Bone char heavy metals water treatment X-ray photoelectron spectroscopy

Introduction

Tailoring the physicochemical properties of adsorbents for water treatment and purification is relevant from an economical and technical point of view (Dastgheib et al., 2004). Process efficacy and operational costs of the adsorption process highly depend on the characteristics of the adsorbent, especially its porous structure, and surface chemistry. The synthesis conditions have a significant impact on the physicochemical properties of the adsorbent and a key issue is to identify the best route for the desired application. However, this is not an easy task because several variables are involved in the synthesis (Rivera-Utrilla et al., 2011). To date, several routes have been reported for the preparation of tailored adsorbents for the removal of both inorganic and organic pollutants, but these studies have mainly focused on activated carbons obtained from lignocellullosic precursors (Bhatnagar et al., 2013; Hokkanen et al., 2016; Rivera-Utrilla et al., 2011). The assessment of alternative materials obtained from other sources and the analysis of their surface characteristics seem mandatory to identify promising adsorbents.

Bone char (BC) is playing an important role as adsorbent in water pollution control because of its outstanding physicochemical properties (Flores-Cano et al., 2016). This material is characterized by a mesoporous structure with a specific surface area from 80 to 120 m²/g and can be obtained via the thermal treatment of different bone wastes (Leyva-Ramos et al., 2010; Moreno-Piraján et al., 2010; Patel et al., 2015). It has been widely used in water defluoridation processes (Medellin-Castillo et al., 2014; Rojas-Mayorga et al., 2013) and, during last decades, new applications have emerged for this adsorbent (Ko et al., 2004; Mendoza-Castillo et al., 2015; Pan et al., 2009; Reynel-Avila et al., 2015, 2016). Several studies have reported the removal of a variety of organic and inorganic toxic compounds from aqueous solutions with BC including, for example, pharmaceuticals (Reynel-Avila et al., 2015), dyes (Reynel-Avila et al., 2016) and heavy metal ions (Ko et al., 2004; Mendoza-Castillo et al., 2015; Pan et al., 2009).

The best synthesis route of BC for water treatment is a matter of debate and different experimental conditions have been reported in the literature, e.g. Moreno-Piraján et al. (2010) and Mwaniki (1992). Despite these discrepancies, it is widely accepted that the temperature and the atmosphere used during the thermal treatment are the most important operating parameters that can affect the final adsorbent properties (Moreno-Piraján et al., 2010; Mwaniki, 1992; Rojas-Mayorga et al., 2013, 2015). It is clear that differences in the synthesis conditions of BC may impact its pollutant uptakes. For instance, commercial BCs show fluoride uptakes from 1 to 4 mg/g (Tovar-Gómez et al., 2013). However, recent studies from our research group have concluded that the fluoride adsorption properties of BC can be improved if an optimum pyrolysis process is used where the final fluoride uptake can be as high as ∼7.0 mg/g (Rojas-Mayorga et al., 2013). Similar trends can be expected to enhance the adsorbent performance for other organic and inorganic water pollutants.

Herein, it is convenient to remark that a standardized and optimized synthesis route for BC with tailored characteristics for the adsorption of heavy metal ions has not been identified, despite the relevance of these pollutants in the context of water treatment. Some authors have shown that the metal uptakes of BC may be higher than those obtained for activated carbons and other adsorbents (Choy and McKay, 2005; Ko et al., 2004; Mendoza-Castillo et al., 2015). However, the discrepancies in the synthesis routes cause a high variability on its final physicochemical properties. The identification of appropriate preparation conditions for this adsorbent is relevant to improve its removal performance and to reduce the production costs.

With this in mind, the aim of this study is to identify the best pyrolysis conditions to synthetize bovine BCs with tailored adsorption properties for heavy metal removal from aqueous solutions. This synthesis route has been analyzed in detail to improve the adsorption of Cd²⁺, Ni², and Zn²⁺ ions at batch reactors. Surface interactions involved in the removal of these metallic cations with BC have been studied using different characterization techniques, given an emphasis in the discussion of XPS analysis.

Methodology

Synthesis of BC and tailoring of its adsorption properties for heavy metal removal

Cow femur bones were utilized as precursor for the synthesis of BC via pyrolysis. The raw precursor was washed using boiling deionized water, sieved and dried. The preparation of BC was performed in a tubular furnace Carbolite Eurotherm CTF 12165/550 with a quartz sample holder under a nitrogen atmosphere (i.e., 400 mL/min). BC samples were obtained at different experimental conditions by adjusting the heating rate, pyrolysis temperature, and thermal treatment time, according to a full factorial design. Experimental design used for the adsorbent synthesis is reported in Table 1, which involved 20 operating conditions. For this experimental design, 27 g of cow femur bones were used as a load in the tubular furnace for the BC synthesis and all adsorbent samples were washed with deionized water and dried before their use in heavy metal adsorption. A mean adsorbent size of ∼0.67 mm was utilized in all experiments. The product yield and adsorption properties of the BC samples for the removal of heavy metals Cd²⁺, Ni², and Zn²⁺ were quantified and used as response variables of the factorial design. The amount of heavy metals adsorbed on the BC was obtained at batch reactors. Adsorption studies were performed by triplicate employing metallic solutions with an initial concentration of 500 mg/L, adsorbent dosage of 2 g/L, equilibrium time of 38 h at 30℃, and pH 5. Results of preliminary experiments showed that these conditions (i.e., adsorbent ratio, pH and temperature) offered the best adsorbent performance. Heavy metals uptakes (q, mg/g) were calculated with the mass balance equation for a batch reactor

q = \frac{({[M^{2 +}]}_{0} - {[M^{2 +}]}_{f}) V_{soln}}{m_{bc}}

(1)

where [M²⁺]₀ and [M²⁺]_f are the initial and final concentrations of heavy metals in the adsorption experiment (mg/L), m_bc is the mass of BC sample used (g), and V_soln is the volume of heavy metal solution (L), respectively. A statistical analysis of the results obtained from the experimental design was carried out to determine the effect of pyrolysis conditions on the heavy metals adsorption properties of BC and the best synthesis route was identified. Statistica® software was used for data analysis.

Table 1.

Metal uptakes and yields of bone char obtained at different pyrolysis conditions.

	Pyrolysis conditions				Metal uptake, mg/g
Sample	Temperature, ℃	Heating rate, ℃/min	Residence time, h	Yield, %	Cd²⁺	Ni²⁺	Zn²⁺
1	650	5	2	75.33	71.1 ± 0.64	67.4 ± 0.85	62.3 ± 0.78
2		5	4	73.45	74.8 ± 0.78	66.2 ± 1.05	61.6 ± 1.08
3		10	2	74.13	73.5 ± 0.59	68.9 ± 0.99	63.0 ± 0.83
4		10	4	73.18	72.4 ± 0.90	69.3 ± 0.96	63.2 ± 0.86
5	700	5	2	75.38	80.1 ± 0.82	74.9 ± 0.91	65.5 ± 0.96
6		5	4	73.29	80.2 ± 0.99	75.4 ± 0.82	64.9 ± 0.90
7		10	2	74.25	81.9 ± 0.79	78.5 ± 0.85	66.7 ± 0.71
8		10	4	73.08	80.7 ± 0.85	77.7 ± 0.79	65.0 ± 0.41
9	800	5	2	75.41	73.2 ± 0.86	70.4 ± 0.93	60.7 ± 0.99
10		5	4	73.24	71.4 ± 0.50	68.2 ± 0.86	61.0 ± 0.57
11		10	2	74.28	72.5 ± 0.42	69.9 ± 0.71	62.7 ± 0.65
12		10	4	73.12	72.4 ± 0.78	67.2 ± 0.58	60.7 ± 0.70
13	900	5	2	73.59	67.9 ± 0.75	57.3 ± 0.51	54.5 ± 0.80
14		5	4	71.79	67.2 ± 0.82	58.8 ± 0.70	55.1 ± 0.91
15		10	2	72.91	68.9 ± 0.79	58.5 ± 0.71	55.2 ± 0.64
16		10	4	71.23	66.7 ± 0.79	57.9 ± 0.85	52.7 ± 0.93
17	1000	5	2	72.73	67.0 ± 0.94	56.5 ± 0.94	50.2 ± 0.86
18		5	4	71.7	68.0 ± 0.73	54.6 ± 0.99	51.2 ± 0.80
19		10	2	71.91	68.0 ± 0.57	56.4 ± 0.92	52.0 ± 0.71
20		10	4	70.53	66.3 ± 0.80	54.5 ± 0.73	51.0 ± 0.78

The best adsorbent was utilized in additional removal experiments for quantifying adsorption kinetics and isotherms of Cd²⁺, Ni², and Zn²⁺. Kinetics studies were performed with an initial concentration of the heavy metals of 500 mg/L and an operating time up to 38 h at pH 5 and 30℃. Adsorption and intraparticle diffusion rates were calculated using the pseudo-second order kinetic model and the Weber-Morris approach, respectively. These models are defined as

q_{t} = \frac{q_{te}^{2} kt}{1 + q_{te} kt}

(2)

q_{t} = k_{d} t^{0.5} + c

(3)

where q_t is the metal uptake at time t, k and k_d are the adsorption and intraparticle diffusion rates, and c is the intercept of Weber-Morris model that is related to the mass transfer across the boundary layer, respectively.

Maximum adsorption capacities were obtained from adsorption isotherms at pH 5 and 30℃ with metal concentrations ranging from 50 to 800 mg/L and an equilibrium time of 38 h. Experimental data were fitted to Langmuir and Sips isotherm models

q_{e} = \frac{q_{max} K_{L} C_{e}}{1 + K_{L} C_{e}}

(4)

q_{e} = \frac{q_{s} a_{s} C_{e}^{n}}{1 + a_{s} C_{e}^{n}}

(5)

where q_e is the equilibrium metal uptake, C_e is the equilibrium metal concentration, K_L is the Langmuir equilibrium constant, q_max is the monolayer adsorption capacity; while q_s, a_s, and n are the adjustable parameters of Sips model, respectively. Results of all adsorption models are reported in Table 2.

Table 2.

Results of kinetic and isotherm data modeling for the adsorption of heavy metals on bone char.

Adsorption model	Parameter	Cd²⁺	Ni²⁺	Zn²⁺
$q_{t} = \frac{q_{te}^{2} kt}{1 + q_{te} kt}$	R ²	0.99	0.98	0.99
	E, %	5.09 ± 3.33	4.71 ± 4.07	1.59 ± 1.26
	k, g/mg h	2.57E−03	2.16E−03	3.72E−03
	q_te, mg/g	88.08	90.56	74.28
	h = $q_{te}^{2}$ k, mg/g h	19.97	17.69	20.54
$q_{t} = k_{d} t^{0.5} + C$	R ²	0.96	0.80	0.89
	E, %	5.53 ± 3.7	15.08 ± 15.3	8.68 ± 9.2
	k_d, mg/g h^0.5	11.29	10.42	8.36
	C	18.54	21.64	22.31
$q_{e} = \frac{q_{max} K_{L} C_{e}}{1 + K_{L} C_{e}}$	R ²	0.75	0.98	0.91
	E, %	22.10 ± 11.2	11.03 ± 10.4	18.30 ± 7.3
	K_L, L/mg	0.17	4.51E−03	0.11
	q_max, mg/g	93.16	95.21	73.28
$q_{e} = \frac{q_{s} a_{s} C_{e}^{n}}{1 + a_{s} C_{e}^{n}}$	R ²	0.98	0.95	0.98
	E, %	9.08 ± 6.6	10.00 ± 4.6	5.59 ± 3.8
	a _s	0.05	3.71E−03	0.11
	n	0.35	1.60	0.45
	q_s, mg/g	455.23	455.34	134.33

Additionally, the multicomponent adsorption of these heavy metals on BC was studied using binary and ternary solutions at pH 5 and 30℃. These tests were performed with the same adsorbent ratio and equilibrium time from the adsorption isotherms. Tables 3 and 4 show the initial metal concentrations used for these multi-metallic adsorption tests. A data analysis was carried out with these multicomponent studies to quantify the impact of co-ion presence on the adsorption behavior of BC.

Table 3.

Multicomponent adsorption of heavy metal ions on bone char using binary mixtures Cd²⁺–Ni²⁺, Zn²⁺–Cd²⁺, and Zn²⁺–Ni²⁺ at pH 5 and 30℃.

M₁–M₂	M₁, mg/L	M₂, mg/L	q_M1, mg/g	q_M2, mg/g
Cd²⁺–Ni²⁺	100	50	39.98	7.90
	250	50	67.96	4.25
	500	50	108.41	2.66
	100	250	37.30	14.32
	250	250	65.78	10.97
	500	250	104.25	6.29
	100	500	32.72	19.76
	250	500	62.32	15.21
	500	500	99.25	7.37
Zn²⁺–Cd²⁺	50	100	17.19	37.17
	300	100	47.37	30.90
	500	100	62.42	24.74
	50	250	13.39	61.71
	300	250	39.17	56.09
	500	250	54.02	49.62
	50	500	7.83	83.52
	300	500	32.02	78.21
	500	500	39.77	72.24
Zn²⁺–Ni²⁺	50	50	32.22	5.36
	300	50	51.06	2.27
	500	50	59.09	0.39
	50	250	23.11	26.54
	300	250	40.99	16.07
	500	250	53.62	8.17
	50	500	18.38	52.34
	300	500	30.54	42.09
	500	500	45.19	33.39

Table 4.

Multicomponent adsorption of heavy metal ions on bone char using ternary mixtures Cd²⁺–Ni²⁺–Zn²⁺ at pH 5 and 30℃.

M₁–M₂–M₃	M₁, mg/L	M₂, mg/L	M₃, mg/L	q_M1, mg/g	q_M2, mg/g	q_M3, mg/g
Cd²⁺–Ni²⁺–Zn²⁺	100	50	50	38.68	4.64	17.18
	100	250	300	33.65	14.81	48.36
	100	500	500	26.90	14.68	63.63
	250	50	300	58.85	1.81	41.30
	250	250	500	51.39	7.83	53.96
	250	500	50	66.49	18.87	13.13
	500	50	500	108.09	1.12	45.55
	500	250	50	98.62	12.34	8.40
	500	500	300	84.26	23.70	30.26

Preparation of metallic solutions and its quantification

All removal experiments were done using metal solutions prepared with reactive grade nitrate salts (JT Baker) and deionized water. Concentrations of heavy metals in aqueous solution were quantified via atomic absorption spectroscopy (ICE 3000 Thermo Scientific) with linear calibration curves.

Study of surface chemistry of BC

Surface chemistry of BC was studied using different characterization techniques. Transmission FTIR spectra were utilized for identifying the surface functional groups with a Bruker IFS 66/S spectrophotometer. All samples were analyzed with spectroscopic grade KBr using a scan time of 200 scans and a resolution of 4 cm⁻¹. The analysis of adsorbent crystalline structure was performed using X-ray diffraction (XRD) measurements in the 10 ≤ 2θ ≤ 80 range, which were obtained with a X-ray diffractometer Bruker D8-Advance equipped with Göebel mirror that has a tube with copper anode RX and radiation Cu Kα (λ = 1.5406 Å). Database from ICDD (International Center for Diffraction Data) was used in this analysis. X-ray photoelectron spectroscopy (XPS) analysis of BC samples was carried out with a K-ALPHA Thermo Scientific spectrometer using Al K-alpha radiation (1486.6 eV), which was monochromatized by a twin crystal monochromator yielding a focused X-ray spot with a diameter of 400 µm at 3 mA × 12 kV. Details of the experimental protocol of the XPS measurements have been reported in a previous study (Rojas-Mayorga et al., 2015). XPS results were analyzed considering a data fitting based on a combination of Lorentzian (30%) and Gaussian (70%) lines. Binding energies were obtained with an accuracy of ±0.2 eV and they were referenced to the C 1s line (i.e., 284.6 eV). The elemental composition was quantified by X-ray fluorescence (XRF) spectroscopy using a sequential X-ray spectrometer (Philips Magix Pro), which was equipped with a rhodium tube and beryllium window. The textural parameters of BC were determined via N₂ adsorption–desorption isotherms at −196℃ with a home-made fully automated equipment designed and constructed by the Advanced Materials group (LMA) of University of Alicante (Spain). Morphology and composition of the BC were obtained by scanning electron microscope/energy-dispersive X-ray spectroscopy (SEM/EDX) with a FE-SEM system (Quanta FEG 650, FEI). Removal mechanism and surface interactions between heavy metal ions and BC were discussed using the results from these characterization techniques.

Results and discussion

Best conditions for the synthesis of BC for heavy metal removal from water and its adsorption properties

Table 1 shows the metal uptake for the twenty samples of BC obtained via pyrolysis. Adsorbent yields ranged from 70.5% to 75.4% and it is clear that the impact of the thermal treatment on this parameter was insignificant. Concerning the adsorption capacity, the metal uptake of tested samples ranged from 50.2 to 66.7 mg/g for Zn²⁺, 54.5 to 78.5 mg/g for Ni²⁺ and 66.3 to 81.9 mg/g for Cd²⁺ ions. Pyrolysis conditions have a significant impact on the metal adsorption properties. The adsorption capacities increased with the pyrolysis temperature from 650℃ to 700℃ and decreased at synthesis temperatures >700℃. Note that this trend prevails for the three heavy metal ions. The highest metal uptakes were obtained at 700℃ being 81.9 ± 0.79 for Cd²⁺, 78.5 ± 0.85 for Ni²⁺ and 66.7 ± 0.71 for Zn²⁺. Results showed that the best pyrolysis conditions correspond to 700℃ during 2 h and a heating rate of 10℃/min. It is interesting to remark that this synthesis route agreed with that reported for BC synthesis using fluoride ion as target pollutant (Rojas-Mayorga et al., 2013). The dehydroxylation process of hydroxyapatite and the decomposition of carbonates of BC occurred at pyrolysis temperatures higher than 700℃ (Rojas-Mayorga et al., 2013). This thermal effect on adsorbent structure may impact its ion exchange properties for heavy metal removal.

Results for kinetic and equilibrium adsorption studies in mono-metallic solutions are reported in Figures 1 –3. Metal adsorption kinetics fitted the pseudo-second order model and the initial adsorption rates ranged from 17.7 to 20.5 mg/g h, while the kinetic constants k varied from 0.00216 to 0.00372 g/mg h. Note that a larger k means a lower adsorption rate (see Table 2). Weber-Morris kinetic analysis showed that the intra-particle diffusion is not the rate limiting stage (Figure 2), which is consistent with the results reported in other studies for heavy metal removal using BC (Mendoza-Castillo et al., 2015). Intraparticle diffusion rates for heavy metal adsorption are reported in Table 2 and they ranged from 0.9 to 11.3 mg/g h^0.5. These results suggested that the surface diffusion has a relevant role as the rate-limiting step in the adsorption of metallic ions on BC.

Figure 1.

Kinetics for the adsorption of Cd²⁺, Ni²⁺, and Zn²⁺ ions on bone char at 30℃ and pH 5.

Figure 2.

Weber-Morris analysis for the kinetics of the adsorption of (a) Cd²⁺, (b) Ni²⁺, and (c) Zn²⁺ ions on bone char at 30℃ and pH 5.

Figure 3.

Isotherms for the adsorption of Cd²⁺, Ni²⁺, and Zn²⁺ ions on bone char at 30℃ and pH 5.

Adsorption isotherms for tested metal ions are given in Figure 3. They can be classed as Langmuir-type isotherm (L2) of Giles classification (Giles et al., 1974), which corresponds to a favorable adsorption process. Results of isotherm data fitting with Langmuir and Sips models are reported in Figure 3 and Table 2. Overall, Sips equation satisfactorily fitted these experimental data but with an overestimation of the monolayer adsorption capacity, see Table 2. The saturation experimental uptakes obtained for this adsorbent ranged from 68.3 to 119.4 mg/g where the removal behavior was: Cd²⁺ > Zn²⁺ > Ni²⁺. These adsorption capacities are higher than those reported for other carbon-based adsorbents used in heavy metal removal (Reynel-Avila et al., 2011). As stated above, previous studies have reported the removal of heavy metal ions with BC, e.g. Ko et al. (2004) and Moreno et al. (2010). Ko et al. (2004) analyzed the adsorption of Cd²⁺ and Zn²⁺ ions on the commercial BC Brimac 216 where the monolayer adsorption capacities were 67.8 and 33.7 mg/g, respectively. On the other hand, Moreno et al. (2010) reported the Ni²⁺ adsorption using a BC obtained from pyrolysis at 800℃ during 5 h where the metal uptake was ∼40 mg/g. These results confirm that the proper selection of the pyrolysis conditions is paramount for obtaining the best adsorbent. In this study, the metal adsorption properties of BC have been improved up to 76%, 71%, and 143% for Cd²⁺, Ni²⁺, and Zn²⁺ ions, respectively. It is also relevant to highlight the outstanding performance of BC for Zn²⁺ adsorption, which is a difficult water pollutant to be removed.

Tables 3 and 4 show the results of multicomponent adsorption of heavy metal ions in both binary and ternary systems using the best BC. These results confirmed the presence of a strong competitive (i.e., antagonistic) adsorption process where the presence of co-ions decreased the metal uptakes. The multicomponent adsorption capacity of BC in binary mixtures for Cd²⁺ ions varied from 24.7 to 108.4 mg/g, while the uptakes of Ni²⁺ and Zn²⁺ ions were 0.4–52.3 mg/g and 7.8–62.4 mg/g, respectively. The impact of co-ion concentration on multicomponent adsorption tests can be determined via the ratio of adsorption capacities Rq_metal = q_metal,mix / q_metal where q_metal,mix is the uptake of BC for metal i in the multicomponent solution and q_metal is the uptake of the same metal using a mono-metallic solution at the same operating conditions (i.e., initial metal concentration, pH and temperature). For binary mixtures, Rq of tested metal ions ranged from 0.03 to 0.94, with Ni²⁺ showing the lowest values. Note that these ratios decreased with increments of the co-ion concentration due to the competition of metallic ions by the binding sites of the adsorbent surface. Cd²⁺ and Ni²⁺ ions have the highest values of electronegativity and the lowest value of hydrated ionic radii, which are cation properties that could favor the selectivity of the ion exchange process involved in the multi-metallic adsorption on BC (Choy et al., 2005). The adsorption capacities in ternary mixtures were 26.9–108.1 mg/g for Cd²⁺, 1.1–23.7 mg/g for Ni²⁺, and 8.4–63.6 mg/g for Zn²⁺ ions, see Table 4. Ni²⁺ uptake was highly impacted by the presence of both Cd²⁺ and Zn²⁺ ions. Values of Rq ranged from 0.09 to 0.95 for the heavy metal ions in the ternary mixture at tested experimental conditions.

Surface chemistry of BC and adsorption mechanism

Figure 4 shows a SEM image and EDX results for the BC sample obtained at 700℃. SEM micrograph displays a rough morphology, non-porous surface without defined geometry, see Figure 4(a). EDX analysis is reported in Figure 4(b) and indicates that the adsorbent is mainly composed of calcium (Ca) and phosphorus (P) where there are also some minor components such as oxygen (O) and carbon (C). Textural parameters of the BC obtained at the best pyrolysis conditions are reported in Figure 5(a). This adsorbent has a Brunauer–Emmett–Teller (BET) surface area of 85 m²/g, a total pore volume of 0.24 cm³/g with a mesopore volume of 0.20 cm³/g, respectively. On the other hand, results of XRD analysis are given in Figure 5(b). XRD patterns showed the characteristic peaks of the known phases of pure hydroxyapatite. The main (h k l) indices reported for the pure hydroxyapatite are (002), (211), (112), (300), (202), (310), (222), (213), and (004), which were identified in the pattern of the BC sample. Note that this result is also consistent with the phases listed in the ICDD database. XRD analysis showed that there were no major changes in the peak position resulting from the metal ion adsorption on BC (i.e., samples labeled as BC-Cd, BC-Ni and BC-Zn in Figure 5b). However, the characteristic reflections of the crystalline phase of the hydroxyapatite partially lost intensity. This result indicated that the hydroxyapatite lost its crystallinity due to the deposition of heavy metal ions on the adsorbent surface. These metallic ions are adsorbed on the hydroxyapatite of BC surface via an ion exchange mechanism with Ca²⁺ ions leading to the formation of a new structure [Ca_10-xM_x(PO₄)₆(OH)₂], where x is the adsorbed heavy metal (Choy and McKay, 2005; Mendoza-Castillo et al., 2015).

Figure 4.

(a) SEM image and (b) EDX results of bone char sample.

Figure 5.

Results of (a) textural parameters, (b) XRD, (c) XRF, (d) FTIR, and (e) XPS of the best bone char obtained via pyrolysis. Label BC is used for raw bone char, while BC–Cd, BC–Ni, and BC–Zn correspond to the adsorbent loaded with Cd²⁺, Ni²⁺, and Zn²⁺ ions, respectively.

The elemental composition (wt %) of the best adsorbent is reported in Figure 5(c). XRF analysis confirmed the chemical composition of hydroxyapatite in BC, which is mainly constituted of calcium and phosphorus, i.e., 39 and 19%, respectively. This analysis also showed the presence of Cd²⁺, Ni²⁺, and Zn²⁺ in the samples BC-Cd, BC-Ni, and BC-Zn where the Ca content decreased in these samples. Previous studies have concluded that divalent metal cations have a strong selectivity towards the Ca²⁺ in the hydroxyapatite lattice via an ion exchange mechanism (Cheung et al., 2000; Rosskopfová et al., 2011; Stötzel et al., 2009).

Figure 5(d) shows the FTIR characteristic absorption bands of the hydroxyapatite and carbonaceous phase of BC at 3430, 2930, 1630, 1460, 1420, 1030, 960, 875, 605, and 565 cm⁻¹ (Rojas-Mayorga et al., 2013, 2015). The stretching vibration of the OH⁻ groups corresponds to the signal at ∼3430 cm⁻¹ (Mondal et al., 2012; Rojas-Mayorga et al., 2013). The asymmetric stretching vibration mode of ${PO}_{4}^{3 -}$ group is given by the signal at ∼1030 cm⁻¹, while the P–O bending vibration mode corresponds to the bands at 960 and 605 cm⁻¹ (Mondal et al., 2012; Rojas-Mayorga et al., 2013). The absorption band at 565 cm⁻¹ can be assigned to the calcium present in the inorganic structure. This band is characteristic of the bond between calcium and the phosphate group (Lurtwitayapont and Srisatit, 2010; Rocha et al., 2005; Rojas-Mayorga et al., 2015; Varma et al., 2005). The band of ${CO}_{3}^{2 -}$ group can be identified at 875 cm⁻¹ (Figueiredo et al., 2010; Lurtwitayapont and Srisatit, 2010; Varma et al., 2005) and the remaining bands are associated to the vibrations of aromatic rings in the carbon phase of BC (2930, 1630, ∼1460 and ∼1420 cm⁻¹) (Kyzas et al., 2013; Rojas-Mayorga et al., 2015). After the heavy metal adsorption, some changes were observed in the FTIR spectra of samples BC-Cd, BC-Ni, and BC-Zn. The Ca–O signal at 565 cm⁻¹ changed and its intensity was lower than that of the spectrum for BC sample. This result is also an evidence of the ion exchange between the heavy metal ions and the calcium from the hydroxyapatite contained in the adsorbent structure.

Figure 5(e) shows the XPS survey spectra of BC samples that contain the photoelectron peaks belonging to oxygen (O), calcium (Ca), phosphorus (P), and carbon (C). These peaks correspond to the hydroxyapatite and the amorphous carbonaceous phase from the adsorbent. Note that the peaks of Cd, Ni, and Zn are identified for samples BC-Cd, BC-Ni, and BC-Zn. A detailed analysis of these XPS results was performed. The Ca 2p peak for raw BC has been deconvoluted into two peaks consisting of Ca 2p_1/2 peak at 351.4 eV and Ca 2p_3/2 peak at 347.9 eV (see Figure 6). These signals indicate the interaction Ca–O. However, the Ca 2p_1/2 peak at 351.4 eV corresponds to the presence of Ca-OH and to the calcium linked to carbonate group (Wang et al., 2014). The main Ca 2p_3/2 peak located at 347.9 eV can be associated to calcium atoms bonded with a phosphate group (Dupraz et al., 1999). The P 2p spectrum has been deconvoluted into two peaks P 2p_1/2 at 134.7 eV and P 2p_3/2 at 133.5 eV (see Figure 7). These energies can be associated with P=O and P–O bonds from phosphate group. The oxygen spectra (O 1s) for BC sample showed two peaks at 534.1 and 531.8 eV, see Figure 8. The O 1s signal at 534.1 eV corresponds to –OH groups, while the peak at 531.8 eV could be attributed to ${PO}_{4}^{3 -}$ groups (Rojas-Mayorga et al., 2015; Wang et al., 2014). These findings are consistent with XPS results reported for Ca 2p, P 2p, and O 1s in the hydroxyapatite (Dupraz et al., 1999; Ohtsu et al., 2013; Wang et al., 2014; Yao et al., 2010).

Figure 6.

Deconvolution of XPS Ca 2p spectra of raw bone char (BC) and metal-loaded bone chars (BC-Cd, BC-Ni and BC-Zn).

Figure 7.

Deconvolution of XPS P 2p spectra of raw bone char (BC) and metal-loaded bone chars (BC-Cd, BC-Ni and BC-Zn).

Figure 8.

Deconvolution of XPS O 1s spectra of raw bone char (BC) and metal-loaded bone chars (BC-Cd, BC-Ni and BC-Zn).

After heavy metal adsorption, several changes in XPS spectra of BC were identified in the region of Ca 2p, P 2p, and O 1s, while new contributions were observed for Cd 3d, Ni 2p, and Zn 2p XPS spectra, see Figure 9. The intensity of binding energies of Ca 2p XPS spectra decreased and a chemical shift to lower energy region was observed for samples BC-Ni, BC-Zn, and BC-Cd. This effect is due to the loss of Ca–O bonds and the change in electron density of remaining Ca atoms in the BC structure after the adsorption process. As stated, this phenomenon is caused by the ion exchange of the calcium from the hydroxyapatite structure of BC by the heavy metals in solution. However, the intensity of this ion exchange process depends on the nature of the heavy metal used in the adsorption experiment. It is convenient to remark that the Ca 2p spectrum of sample BC-Cd showed the major change indicating that cadmium removal caused the highest calcium exchange. These findings are consistent with the BC adsorption isotherms, which showed the highest uptakes for cadmium. The new Ca 2p binding energy identified for samples BC-Cd, BC-Ni, and BC-Zn corresponds to Ca–O interactions from hydroxyapatite. These results also confirmed that the affinity of BC was different for the tested metal ions. For illustration, the elemental composition (at %) of different BC samples is reported in Table 5.

Figure 9.

Deconvolution of XPS cadmium, nickel, and zinc spectra of metal-loaded bone chars.

Table 5.

Atomic ratios (%) obtained from XPS analysis for bone char samples before and after heavy metal adsorption.

Bone char sample	Atomic ratio, %
Bone char sample	C	P	Ca	O	Cd	Ni	Zn
BC	20.46	12.75	18.82	47.97	–	–	–
BC–Cd	23.65	12.20	14.92	46.08	3.15	–	–
BC–Ni	23.39	12.13	16.58	46.83	–	1.07	–
BC–Zn	22.46	11.52	15.01	48.58	–	–	2.43

XPS spectra of samples BC-Ni and BC-Cd showed the largest change in the P 2p region. These spectra get wider after the removal process, thus revealing the presence of additional contributions to the phosphate group. These new binding energies correspond to P–O and P=O bonds of the phosphate group from hydroxyapatite. These contributions may be associated with the interaction of Cd and Ni with the phosphate group, due to the aforementioned displacement of the calcium atoms from BC structure by these ions. The O 1s energy level for the initial contribution of the BC at 534.1 eV shifted to lower energy values of 533.1, 533.4, and 533.7 eV in the samples BC-Cd, BC-Ni, and BC-Zn, respectively. This is due to the displacement of calcium atoms from hydroxyapatite by heavy metal atoms during adsorption on BC leading to the formation of new metal–OH interactions (Su et al., 2013). The peak at 531.8 eV in sample BC also shifted to lower energy values: 531.2, 531.5, and 531.6 eV for samples BC-Cd, BC-Ni, and BC-Zn, respectively. This change could be caused by the interaction of the phosphate groups with the heavy metal ions.

Figure 9(a) shows the Cd 3d energy level that has been deconvoluted into two Cd 3d_5/2 peaks at 407.0 eV and 405.4 eV. Both binding energies are associated with the interactions Cd–O (Haemers et al., 1984) and, according to O 1s and P 2p energy levels, these contributions could be the interaction of Cd ions with hydroxyl and phosphate groups. Ni 2p spectrum has been deconvoluted into four peaks at 865.45, 862.05, 858.29, and 856.29 eV (see Figure 9b). These signals can be associated with the Ni–O interactions such Ni–phosphate, Ni-OH and nickel oxides with different electronic density (Nesbitt et al., 2000). The Zn 2p XPS spectra showed a single signal at 1022.6 eV that corresponds to the Zn–O bond, which is formed between the oxygen from hydroxyapatite surface and the adsorbed Zn²⁺ ions (see Figure 9c). This interaction can correspond to Zn-OH (Haemers et al., 1984).

These characterization results confirmed that the adsorption of heavy metals on BC occurred via an ion exchange process between the structural calcium from hydroxyapatite and the heavy metal ions in solution (Mendoza-Castillo et al., 2015)

\equiv C a^{2 +} + M^{2 +} = \equiv M^{2 +} + C a^{2 +}

(6)

where M²⁺ is the metal ion in aqueous solution and ≡ symbolizes the BC surface, respectively. This mechanism also explains the antagonistic multicomponent adsorption of tested heavy metals in binary and ternary systems.

Conclusions

A tailored BC for heavy metal adsorption from aqueous solution has been obtained via an optimized pyrolysis process. Results showed that a pyrolysis temperature >700℃ has a significant impact on metal uptakes of this adsorbent due to the dehydroxylation of hydroxyapatite and the decomposition of carbonates from BC. The best synthesis route improved the heavy metal adsorption properties of BC up to 143% with respect to the results reported in literature for other commercial adsorbents. BC showed an outstanding performance for the adsorption of Zn²⁺ ions, which is a relevant and persistent water pollutant. However, the multicomponent removal of Zn²⁺, Cd²⁺, and Ni²⁺ ions in both binary and ternary mixtures is a strong antagonistic adsorption process. XPS characterization analysis confirmed that the main mechanism involved in heavy metal adsorption with BC is an ion exchange process where metal–oxygen interactions from hydroxyapatite could play a relevant role. In summary, this study reports new insights to improve the surface chemistry of BC for water treatment and purification.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Appendix

References

Bhatnagar

Hogland

Marques

(2013) An overview of the modification methods of activated carbon for its water treatment applications. Chemical Engineering Journal 219: 499–511.

Cheung

Porter

McKay

(2000) Sorption kinetics for the removal of copper and zinc from effluents using bone char. Separation and Purification Technology 19: 55–64.

Choy

KKH

McKay

(2005) Sorption of metal ions from aqueous solution using bone char. Environment International 31: 845–854.

Dastgheib

Karanfil

Cheng

(2004) Tailoring activated carbons for enhanced removal of natural organic matter from natural waters. Carbon 42: 547–557.

Dupraz

Nguyen

Richard

(1999) Influence of a cellulosic ether carrier on the structure of biphasic calcium phosphate ceramic particles in an injectable composite material. Biomaterials 20: 663–673.

Figueiredo

Fernando

Martins

(2010) Effect of the calcination temperature on the composition and microstructure of hydroxyapatite derived from human and animal bone. Ceramics International 36: 2383–2393.

Flores-Cano

Leyva-Ramos

Carrasco-Marin

(2016) Adsorption mechanism of Chromium (III) from water solution on bone char: Effect of operating conditions. Adsorption 22: 297–308.

Giles

Smith

Huitson

(1974) A general treatment and classification of the solute adsorption isotherm. I. Theoretical. Journal of Colloid and Interface Science 47: 755–765.

Haemers

Verbist

Maroie

(1984) Surface oxidation of polycrystalline α(75%Cu/25%Zn) and β(53%Cu/47%Zn) brass as studied by XPS: Influence of oxygen pressure. Applications of Surface Science 17: 463–467.

10.

Hokkanen

Bhatnagar

Sillanpaa

(2016) A review on modification methods to cellulose-based adsorbents to improve adsorption capacity. Water Research 91: 156–173.

11.

DCK

Cheung

Choy

KKH

(2004) Sorption equilibria of metal ions on bone char. Chemosphere 54: 273–281.

12.

Kyzas

Lazaridis

Deliyanni

(2013) Oxidation time effect of activated carbons for drug adsorption. Chemical Engineering Journal 234: 491–499.

13.

Leyva-Ramos

Rivera-Utrilla

Medellin-Castillo

(2010) Kinetic modeling of fluoride adsorption from aqueous solution onto bone char. Chemical Engineering Journal 158: 458–467.

14.

Lurtwitayapont

Srisatit

(2010) Comparison of lead removal by various types of swine bone adsorbents. Environment Asia 3: 32–38.

15.

Medellin-Castillo

Leyva-Ramos

Padilla-Ortega

(2014) Adsorption capacity of bone char for removing fluoride from water solution. Role of hydroxyapatite content, adsorption mechanism and competing anions. Journal of Industrial and Engineering Chemistry 20: 4014–4021.

16.

Mendoza-Castillo

Bonilla-Petriciolet

Jáuregui-Rincón

(2015) On the importance of surface chemistry and composition of bone char for the sorption of heavy metals from aqueous solution. Desalination and Water Treatment 54: 1651–1662.

17.

Mondal

Dey

(2012) Studies on processing and characterization of hydroxyapatite biomaterials from different bio wastes. Journal of Minerals and Materials Characterization and Engineering 11: 55–67.

18.

Moreno

Gómez

Giraldo

(2010) Removal of Mn, Fe, Ni and Cu ions from wastewater using cow bone charcoal. Materials 3: 452–466.

19.

Moreno-Piraján

Gómez-Cruz

García-Cuello

(2010) Binary system Cu(II)/Pb(II) adsorption on activated carbon obtained by pyrolysis of cow bone study. Journal of Analytical and Applied Pyrolysis 89: 122–128.

20.

Mwaniki

(1992) Fluoride sorption characteristics of different grades of bone charcoal based on batch tests. Journal of Dental Research 71: 1310–1315.

21.

Nesbitt

Legrand

Bancroft

(2000) Interpretation of Ni2p XPS spectra of Ni conductors, and Ni insulators. Physics and Chemistry of Minerals 27: 357–366.

22.

Ohtsu

Hiromoto

Yamane

(2013) Chemical and crystallographic characterizations of hydroxyapatite and octacalcium phosphate coatings on magnesium synthesized by chemical solution deposition using XPS and XRD. Surface and Coatings Technology 218: 114–118.

23.

Pan

Wang

Zhang

(2009) Sorption of cobalt to bone char: Kinetics, competitive sorption and mechanism. Desalination 249: 609–614.

24.

Patel

Han

Qiu

(2015) Synthesis and characterisation of mesoporous bone char obtained by pyrolysis of animal bones for environmental application. Journal of Environmental Chemical Engineering 3: 2368–2377.

25.

Reynel-Avila

Mendoza-Castillo

Bonilla-Petriciolet

(2015) Assessment of naproxen adsorption on bone char in aqueous solutions using batch and fixed-bed processes. Journal of Molecular Liquids 209: 187–195.

26.

Reynel-Avila

Mendoza-Castillo

Bonilla-Petriciolet

(2016) Relevance of anionic dye properties on water decolorization performance using bone char: Adsorption kinetics, isotherms and breakthrough curves. Journal of Molecular Liquids 219: 425–434.

27.

Reynel-Avila

Mendoza-Castillo

Hernández-Montoya

(2011) Multicomponent removal of heavy metals from aqueous solution using low-cost sorbents. In: Antizar-Ladislao

Sheikholeslami

(eds) Water Production and Wastewater Treatment, New York: Editorial Nova Science Publisher, pp. 69–99.

28.

Rivera-Utrilla

Sánchez-Polo

Gómez-Serrano

(2011) Activated carbon modifications to enhance its water treatment applications. An overview. Journal of Hazardous Materials 187: 1–23.

29.

Rocha

JHG

Lemos

Kannan

(2005) Hydroxyapatite scaffolds hydrothermally grown from aragonitic cuttlefish bones. Journal of Materials Chemistry 15: 5007–5011.

30.

Rojas-Mayorga

Bonilla-Petriciolet

Aguayo-Villarreal

(2013) Optimization of pyrolysis conditions and adsorption properties of bone char for fluoride removal from water. Journal of Analytical and Applied Pyrolysis 104: 10–18.

31.

Rojas-Mayorga

Bonilla-Petriciolet

Silvestre-Albero

(2015) Physico-chemical characterization of metal-doped bone chars and their adsorption behavior for water defluoridation. Applied Surface Science 355: 748–760.

32.

Rojas-Mayorga

Silvestre-Albero

Aguayo-Villareal

(2015) A new synthesis route for bone chars using CO₂ atmosphere and their application as fluoride adsorbents. Microporous and Mesoporous Materials 209: 38–44.

33.

Rosskopfová

Galamboš

Rajec

(2011) Study of sorption processes of strontium on the synthetic hydroxyapatite. Journal of Radioanalytical and Nuclear Chemistry 287: 715–722.

34.

Stötzel

Müller

Reinert

(2009) Ion adsorption behaviour of hydroxyapatite with different crystallinities. Colloids and Surfaces B: Biointerfaces 74: 91–95.

35.

Ramprasad

Han

(2013) Dense CdS thin films on fluorine-doped tin oxide coated glass by high-rate microreactor-assisted solution deposition. Thin Solid Films 532: 16–21.

36.

Tovar-Gómez

Moreno-Virgen

Dena-Aguilar

(2013) Modeling of fixed-bed adsorption of fluoride on bon echar using a hybrid neural network approach. Chemical Engineering Journal 228: 1098–1109.

37.

Varma

Babu

(2005) Synthesis of calcium phosphate bioceramics by citrate gel pyrolysis method. Ceramics International 31: 109–114.

38.

Wang

Zhu

Wang

(2014) Formation mechanism of Ca-deficient hydroxyapatite coating on Mg–Zn–Ca alloy for orthopaedic implant. Applied Surface Science 307: 92–100.

39.

Yao

Ivanisenko

Diemant

(2010) Synthesis and properties of hydroxyapatite-containing porous titania coating on ultrafine-grained titanium by micro-arc oxidation. Acta Biomaterialia 6: 2816–2825.

Tailoring the adsorption behavior of bone char for heavy metal removal from aqueous solution

Abstract

Keywords

Introduction

Methodology

Synthesis of BC and tailoring of its adsorption properties for heavy metal removal

Preparation of metallic solutions and its quantification

Study of surface chemistry of BC

Results and discussion

Best conditions for the synthesis of BC for heavy metal removal from water and its adsorption properties

Surface chemistry of BC and adsorption mechanism

Conclusions

Footnotes

Declaration of Conflicting Interests

Funding

Appendix

References