Comparative study of dry-mixing and wet-mixing activated carbons prepared from waste printed circuit boards by NaOH activation

Abstract

Two activated carbons have been produced from the non-metallic fraction (NMF) of waste printed circuit boards (WPCB) by sodium hydroxide activation by two methods, dry-mixing and wet-mixing. Dry-mixing is the waterless way by direct powder maceration; activated carbon prepared through this way is called AC. Wet-mixing is by impregnation within an aqueous solution; activated carbon prepared through this way is called WAC. Activated carbons differed in physical structure and chemical properties as derived from N₂ adsorption/desorption isotherms and Fourier-transform infrared spectroscopy (FTIR). Batch sorption studies were performed to compare the ciprofloxacin (CIP) adsorption properties of the two carbons. Experiments proved that the method of mixing has great influence on the preparation of activated carbons. AC showed better characteristics and adsorption properties than WAC. AC had 2304 m² g⁻¹ surface area, while WAC had 1141 m² g⁻¹ surface area. Fourier transform infrared spectra indicated the two activated carbons had similar functional groups. For both adsorbents, the adsorption kinetics followed the pseudo-second-order model. The adsorption equilibrium data were very well represented by Langmuir equations.

---

1. Introduction

With the phenomenal advancement and growth in electronic industries, higher demand for electric and electronic equipment (EEE) has made the replacement of electronic equipment more frequent, resulting in large amounts of electronic waste (E-waste) to be disposed of.¹ A current estimate shows nearly 45 million tons of E-waste is generated globally per annum and the number is growing at an exponential rate.²

Waste printed circuit boards (WPCB) are one of the most complicated constituents of WEEE. WPCB, which contains valuable resources together with some hazardous materials, are thus considered to be an attractive secondary resource and an environmental contaminant.³ In order to recycle the valuable resources, hydrometallurgical process which includes various steps of leaching, separation and purification is the main route.^4,5 Unfortunately, toxic water would be generated during these procedures. Novel approaches to reuse WPCB are urgently to be proposed. Because of the high content of organic compounds in non-metallic fraction (NMF) of WPCB,⁶ it can be used as a low-cost raw material for preparing activated carbons. Researches focused on the preparation of activated carbon from WPCB are few.^7,8

Activated carbon can be prepared by different activating agents. The use of alkali hydroxides (NaOH, KOH) for the production of microporous activated carbons has attracted great interest due to the valuable properties of the materials produced by this process. Compared with KOH activations, NaOH activation has advantages such as: (I) lower dosage; (II) cheaper; (III) more environmentally; (IV) less corrosive.⁹ Another reason is that NaOH occurs carbonation in the presence of water. Different mixing ways may display different reactions, which has great influence on the preparation of activated carbon. For the reason above, we choose NaOH as the activating agents.

Ciprofloxacin (CIP) is a synthetic antibiotic widely used for the treatment of several bacterial infections that can end up into water courses due to incomplete metabolism in humans or coming from effluents of drug manufacturers. The presence of CIP in water sources can lead to the development of antibiotic resistant bacteria, which is extremely harmful to the health of the human.¹⁰ Thus, removing CIP from water has become an increasingly important subject. Various methods including adsorption,¹¹ degradation¹² and oxidation¹³ had been studied for removing CIP. Adsorption is considered the most suitable treatment because it inhibits the toxic properties and restricts the transportation into water systems.¹⁴

This study focused on comparison of dry-mixing and wet-mixing to prepare NMF-based activated carbons by NaOH activation. Dry-mixing is the waterless way by direct powder maceration, activated carbon prepared through this way is called AC. Wet-mixing is by impregnation within an aqueous solution, activated carbon prepared through this way is called WAC. The two activated carbons differed with the physical structure and chemical properties which were derived from N₂ adsorption/desorption isotherms and Fourier-transform infrared spectroscopy (FTIR). Batched sorption studies were performed to compare the CIP adsorption properties of AC and WAC. Experiments proved that the method of mixing had great influence on the preparation of activated carbons, and AC showed better characteristics and adsorption properties than WAC.

2. Materials and methods

2.1. Materials

The non-metallic powders were provided by Shandong Zhonglv Eco-recycle Co., Ltd. The properties of the raw materials were represented in Fig. 1. All chemical reagents used in this study were analytical grade. CIP (C₁₇H₁₈FN₃O₃) was purchased from Sangon Biotech (Shanghai, China) Co., Ltd.

Fig. 1 The properties of non-metallic powders.

2.2. Preparation of AC and WAC

First, a carbon-rich carbonized sample was prepared from the pyrolysis of the raw materials under N₂ atmosphere at 600 °C for 2 h. The carbonized sample was solid powder with size lower than 0.25 mm (60 mesh) diameter. Next, the carbonized sample was mixed with NaOH by two different processes: dry-mixing or wet-mixing. The progress flow charts had been showed in Fig. 2.

Fig. 2 The progress flow charts of AC and WAC.

In the dry-mixing, sodium hydroxide powders are directly mixed with a given amount of the carbonized sample. It should be emphasized that the process is done in absence of water.

In the wet-mixing, the carbonized sample was mixed with sodium hydroxide adding 10 mL distilled water. Then stirring often to make sure the sodium hydroxide dissolved completely. After the impregnation process, the mixtures were dried at 110 °C for 5 h.

For the two preparation methods, dry-mixing or wet-mixing, the NaOH/carbonized sample weight ratio was kept equal to 3 : 1. The activation heat-treatments were conducted under N₂ flow at 800 °C for 1 h in a cylindrical furnace. After the heat treatment, a washing process was carried out by using a 2 M HCl solution firstly, and afterwards washed with distilled water until the pH value reached to neutral. Then, carbons were dried at 110 °C for 4 h, ground and sieved by sieves to 0.147 mm. The carbon prepared by dry-mixing was named as AC. The carbon prepared by wet-mixing was named as WAC.

2.3. Characterization methods

SDT Q600 equipment was used to obtain thermo gravimetric analysis (TGA) and differential scanning calorimetry (DSC) diagram. Scanning electron microscope (SEM, Hitachi S-520) was used to study the changes in surface morphology. The BET surface area (S_BET) and porous properties of AC and WAC were determined by N₂ adsorption/desorption isotherms at 77 K using a surface area analyzer (Quantachrome Corporation, USA). The S_BET and the total pore volume (V_tot) were obtained by the manufacturer's software. The S_BET was derived from the Brunauer–Emmett–Teller (BET) theory. The micropore surface area (S_mic), external surface area (S_ext) and the micropore volume (V_mic) were evaluated by the t-plot method. The pore size distribution was obtained by the Density Functional Theory (DFT) method. The mean pore diameter (D_p) was obtained from D_p = 4V_tot/S_BET. The surface functional groups of AC and WAC were recorded by the KBr technique using a Fourier transform infrared (FTIR) spectrometer (Fourier-380FT-IR, USA), and the spectra were recorded in the spectral range 4000–400 cm⁻¹.

2.4. Adsorption procedure

Parallel series of batch adsorption tests were carried out in conical flask with cap containing AC or WAC. In both of them, 0.05 g of solid adsorbent was introduced to 50 mL solution containing a specific amount of CIP (200–800 mg L⁻¹). All pH measurements were carried out using a pH meter (Model pHS-3C, Shanghai, China). The initial pH levels of the experimental solutions were adjusted to constant values by adding various concentrated HCl or NaOH solutions. The flasks were then transferred to a water-bath thermostatic shaker (SHA-B, Shanghai, China) at determined temperature and shook at 170 rpm for 24 h to ensure that the adsorption process reached equilibrium. One conical flask containing 50 mL CIP solution sample was taken off at various time intervals to determine adsorption kinetics. Samples were filtered using a 0.45 μm membrane filter, and the CIP concentration of the filtrate was analyzed using UV/Vis spectrophotometer (UV-754, Shanghai) at the maximum wavelength of 275 nm.

2.5. Adsorption isotherm and kinetic models

In order to explain the adsorption mechanism, the kinetic data were analyzed with pseudo-first-order,¹⁵ pseudo-second-order,¹⁶ Elovich¹⁷ and intra-particle diffusion models.¹⁸ Similarly, the isothermal data were analyzed with Langmuir,¹⁷ Freundlich¹⁹ and Temkin models.²⁰ The kinetic and isotherm models were listed in Table 1. The determination coefficient (R²) and normalized standard deviation (Δq_e) were evaluated to define which models best describes the experimental data. Δq_e was calculated using the following equation:²⁰where N is the number of experimental data points.

Table 1 The kinetic models and isothermal models^a

Kinetic models	Names
a qt – adsorbed amount at any time; qe – adsorbed amount at equilibrium; k1 – rate constant of pseudo-first order model; k2 – rate constant of pseudo-second order model; α and β – Elovich constants; kpi and Ci – intra-particle diffusion constant; qm – maximum adsorption capacity; KL – Langmuir constant; KF and n – Freundlich constants; BT and KT – Temkin constants.
qt = qe(1 − e^−k1t)	Pseudo-first-order
	Pseudo-second-order
	Elovich
qt = kpit^1/2 + Ci	Intra-particle diffusion

---

Isothermal models	Names
	Langmuir
	Freundlich
qe = BT ln(KTce)	Temkin

---

3. Results and discussion

3.1. Thermal analysis

The two samples were prepared by dry-mixing or wet-mixing, respectively. Both of them were heated between 20 and 800 °C under N₂ atmosphere. The thermal analyses are illustrated in Fig. 3. The pyrolysis process can be divided into three steps: an initial weight loss due to the evaporation of moisture ((a): 5.19%; (b): 12.58%), a significant mass loss in 200–500 °C, which was the main weightlessness area for the precursors ((a): 27.86%; (b): 22.31%), benzene reacted with NaOH to form new structures resulting in the loss of organic compounds. A slow mass loss at the last stage between 500 and 800 °C ((a): 8.43%; (b): 6.49%), confirming the stable structure of the resultant char was formed in this temperature range. Because of the different mixing method, the thermal analyses of two samples in the three mass loss stages have significant differences, especially in the first stage, since the precursor of WAC have higher moisture content that the precursor of AC.

Fig. 3 Thermal analysis (DSC/TGA) of dry-mixing (a) and wet-mixing (b).

The carbon-rich carbonized sample was prepared from the pyrolysis of the non-metallic powders under N₂ atmosphere at 600 °C for 2 h. The main reaction process was showed in Fig. 4. The first step of the decomposition process is the release of water and the formation of unsaturated structure. For the structure I, when C–O and C–Br were broken, bisphenol A (BPA) would be generated. For the structure II, the extraction of H resulted in the formation of phenol. Structure II reacted with NaOH under different conditions. The main reactions were showed as follows:²¹

Fig. 4 The pyrolysis of epoxy resin monomer (I: the brominated epoxy resin monomer; II: X is H or Br).

3.2 SEM analysis

The SEM profiles of AC and WAC were presented in Fig. 5. Both the samples had a significant difference in appearance. The fragmentation degree of AC was higher than that of WAC. The activating agents-NaOH played a bigger role in the dry-mixing than in the wet-mixing. During the wet-mixing, part of NaOH occurred carbonation, which prevent the development of porosity. The porosity of two samples was related to the volatile matter and gas generated in the activation progress.

Fig. 5 SEM micrographs of (a) AC and (b) WAC.

3.3 Textural structure

The N₂ adsorption/desorption isotherms at 77 K has been determined in AC and WAC to assess their porous texture, which conclude pore volume, pore area and average pore diameter. Fig. 6 and 7 show the N₂ adsorption/desorption isotherms and pore size distributions, respectively. The structural parameters of AC and WAC were summarized in Table 2.

Fig. 6 The adsorption–desorption isotherms of N₂ at 77 K for AC and WAC.

Fig. 7 Pore size distributions for AC and WAC.

Table 2 Porous structure parameters of AC and WAC

Type	SBET (m^2 g^−1)	Smic (m^2 g^−1)	Smic/SBET (%)	Sext (m^2 g^−1)	Sext/SBET (%)	Vtot (cm^3 g^−1)	Vmic (cm^3 g^−1)	Vmic/Vtot (%)	Vext (cm^3 g^−1)	Vext/Vtot (%)	Dp (nm)	Yield (%)
AC	2304	1938	84.11	366	15.89	1.171	0.847	72.33	0.324	27.67	2.032	34.28
WAC	1141	1073	94.04	68	5.96	0.576	0.474	82.29	0.102	17.71	2.031	38.40

---

According to the classification of International Union of Pure and Applied Chemistry,²² the majority of physisorption isotherms can be grouped into the six types corresponding to six different types of pore structure. From Fig. 6, it can be seen that the N₂ adsorption/desorption isotherms for both adsorbents belongs to types I, corroborating that AC and WAC are microporous carbons. Additionally, the plateau showed nearly parallel to the axis of P/P₀, confirming the presence of micropores with no or few mesopores.²³ Compared with WAC, the pronounced increase in the volume of N₂ adsorbed and desorbed amount at the low pressure region of AC demonstrates the development of microporosity. In addition, the pore size distribution (Fig. 7) of AC and WAC showed that a vast majority of the pores fall into the range of microspore.

As shown in Table 2, WAC showed a low S_BET in comparison to AC, probably because the amount of NaOH was not sufficient react with chars. When the materials containing NaOH and water are in air, thus, in a CO₂ atmosphere, the carbonation of the NaOH occurs (R.1), resulting in the reduction of NaOH. In the conditions of activation experiments (activation temperature of 800 °C), the Na₂CO₃ is stable, is does not turn to any other sodium compound active for the activation, and therefore the carbonation of the NaOH prevent the development of porosity. In another indication that the yield of WAC is higher than that of AC.²⁴

The results above showed that AC has better porous characteristic than WAC. The drying stage during the preparation of WAC is negative, lowering the development of porosity.²⁵

2NaOH + CO₂ ↔ Na₂CO₃ + H₂O (R.1)

3.4 Surface chemistry analysis

The adsorption abilities had much to do with the surface chemistry characteristic. In order to understand the surface chemistry differences of AC and WAC, Table 3 summarizes the pH_pzc and number of surface functional group. The two adsorbents have different values of surface acidity and basicity, which is represented as the amount of acidic groups and basic groups, respectively. AC has higher pH_pzc value, smaller amount of acidic functional group and larger amount of basic functional group than WAC.

Table 3 The surface chemistry of AC and WAC

Acidic and basic function group concentration (mmol g^−1)
Samples	pHpzc	Basic	Lactone	Carboxyl	Phenolic	Total acidic
AC	6.98	0.971	0.580	0.423	0.969	1.972
WAC	6.25	0.722	0.745	0.311	1.053	2.109

---

The FTIR spectrums of AC and WAC revealed different functional groups (Fig. 8). The broad band at 3402 cm⁻¹, which could be attributed to the O–H stretching vibration from OH⁻ groups (characteristic of NaOH).²³ The peak at 1565 cm⁻¹ is attributed to the formation of oxygen functional groups such as C=O stretching in quinine structure.²⁶ The bands at 1402 cm⁻¹ are assigned to C–H bending in alkyl groups. The broad band between 1200 and 1000 cm⁻¹ has been assigned to C–O stretching in alcohols and phenols.²⁶ This part represented various configurations of C, O, H and N bonds in the carbon structures.²⁷

Fig. 8 FTIR spectra of AC and WAC.

Seen from Fig. 8, AC and WAC had very similar spectra, which confirmed that the two samples had the similar functional groups on the surface. It indicated that the activation mechanism of the two preparation methods were the same. Compared with AC, the intensity of the peaks at 3402 cm⁻¹ in WAC reduced. In the initial reaction period for preparing WAC, the carbonation of the NaOH occurs, NaOH reacted with CO₂ to form Na₂CO₃, which weakened the intensity of the characteristic band of NaOH at 3402 cm⁻¹.^25,28

3.5 Adsorption experiments

3.5.1 The effect of contact time and adsorption kinetics of CIP onto AC and WAC. The time required to reach the equilibrium in the adsorption study has been considered as a significant parameter. Fig. 9 illustrated the effects of contact time on the adsorption of CIP onto AC and WAC at 25 °C. The CIP concentration decreased rapidly in the first 4 h, and then the equilibrium was gradually obtained within 13 h. AC exhibit much better sorption and stronger adsorption affinity to CIP due to the larger surface area and total volume compared to WAC. This phenomenon implied that although adsorbent surface area and porous properties were the main affecting factors that affect the adsorption of CIP, other factor like the surface chemistry of adsorbents also may affect the adsorption.

Fig. 9 Effect of contact time on CIP adsorption by AC and WAC.

The kinetic parameters were listed in Table 4. It can be seen the models of pseudo-first-order and Elovich were not suitable to describe the experimental data, presenting the lower R² (0.92–0.98) and the higher Δq_e values (3.05–10.29%). Instead, the pseudo-second-order model was well fitted to the experimental data with a higher R² (>0.99) and lower Δq_e value (<1.5%). The pseudo-second-order model assumed that the adsorption progress was affected by chemical interactions which lead to binding of the adsorbate to the adsorbent surface as strong as covalent bonding, which was consistent with the cation exchange and electrostatic attraction mechanism.¹⁸

Table 4 Kinetic parameters for the adsorption of CIP onto AC and WAC

Kinetic models	Parameters	Adsorbent
		AC	WAC
Pseudo-first-order	qe,exp	550.72	375.74
	qe,cal	394.26	195.02
	k1	0.0043	0.0031
	R^2	0.9612	0.9267
	Δqe	6.08	10.29
Pseudo-second-order	qe,cal	559.91	380.22
	k2	0.0024	0.0003
	R^2	0.9948	0.9986
	Δqe	1.35	1.08
Elovich	α	437.55	286.78
	β	0.0234	0.0137
	R^2	0.9852	0.9790
	Δqe	3.05	4.14
Intra-particle diffusion	kp1	31.83	19.15
	C1	112.3	117.2
	R^2	0.9921	0.9770
	kp2	11.79	4.957
	C2	286.3	236
	R^2	0.9381	0.9515
	kp3	0.002	1.214
	C3	550.6	331.2
	R^2	0.9592	0.8296

---

The intra-particle diffusion model assumed the adsorption mechanism occurred through the diffusion of the adsorbate molecules into the pores of adsorbent material.¹⁶ The plots obtained for q_t versus t^1/2 were shown in Fig. 10. The parameters (k_pi, C_i and R²) were listed in Table 4. As can be seen in Fig. 10, all the adsorption data exhibit multi-linear plots, illustrating that two or more steps influenced the adsorption process. The first linear sections did not pass through the origin, which meant the intra-particle diffusion was not the only rate-controlling step.²⁹ The first stage indicated external mass transfer or external film resistance controlled the CIP adsorption. The second stage described the gradual adsorption, where the intra-particle diffusion was the rate-limiting step. The third stage was assigned to the equilibrium, where the intra-particle diffusion started to slow down due to the extremely low concentration of CIP remaining in solution, resulting in lower k_pi values for both stages.³⁰

Fig. 10 Intra-particle diffusion model for adsorption of CIP onto AC and WAC.

3.5.2 Effect of pH. The adsorption of CIP onto AC and WAC as a function of pH had been shown in Fig. 11. The adsorption of CIP onto AC and WAC was pH-dependent. The main reason for the change of adsorption capacity is the electrostatic interactions between adsorbent and adsorbate.³¹ The molecular structure of CIP is also an important factor to affect the adsorption. CIP has three species (CIP⁻, CIP⁰, CIP⁺) under different pH condition. CIP⁺ is the cationic form existing in the solution pH < 6.1, which is due to the protonation of the amine group. CIP⁰ is the zwitterionic form existing in the solution pH between 6.1 (pk_a1) and 8.7 (pk_a2), which contributes to balance the charge in the solution. CIP⁻ is the anionic form existing in the solution pH > 8.7, which is due to a proton loss from the carboxylic group.³² The adsorption behavior should also be combined with the surface charge characteristics of AC and WAC. The pH_pzc value of the two adsorbents were shown in Table 3. The positively charged adsorbents gradually become negatively charged as the pH value increases.³³ At solution pH < pk_a1 (pk_a1 = 6.1), the CIP molecules and adsorbents surface have opposite charges, thus adsorption is expected to be enhanced. When pH > pk_a2 (pk_a2 = 8.7), the amounts of CIP adsorbed onto adsorbents are expected to be depressed, which was because CIP and adsorbents surfaces were negatively charged and repelled each other.³⁴ The electrostatic effect for the sorption of CIP was shown in Fig. 12.

Fig. 11 The influence of solution pH on the adsorption of CIP onto AC and WAC.

Fig. 12 The electrostatic effect for the sorption of CIP.

3.5.3 Adsorption isotherm. Adsorption isotherms reflect the capacity of the adsorbents and the interaction between adsorbents and adsorbates. The parameters of isotherm models for Langmuir, Freundlich and Temkin were shown in Table 5. The Langmuir model described an adsorption process that occurred upon a homogeneous surface where the adsorbate is distributed in monolayers. According to the Table 5, Langmuir model showed higher determination coefficient and lower normalized standard deviation than Freundlich and Temkin models, which indicating this adsorption process was monolayer coverage.

Table 5 Isotherm parameters for CIP adsorption onto AC and WAC

Isotherm models	Parameters	AC			WAC
		298 K	308 K	318 K	298 K	308 K	318 K
Langmuir	qm	572.11	586.29	603.84	381.29	406.72	429.65
	KL	0.048	0.051	0.059	0.0025	0.0029	0.0035
	R^2	0.9963	0.9903	0.9969	0.9924	0.9942	0.9987
	Δqe	0.89	1.17	0.87	1.09	0.93	0.75
Freundlich	KL	297.10	301.72	315.94	176.67	184.99	195.26
	n	8.63	7.33	9.04	10.77	11.35	12.07
	R^2	0.9814	0.9774	0.9806	0.9733	0.9759	0.9853
	Δqe	1.85	2.26	1.90	2.36	2.30	1.54
Temkin	KT	9.874	14.098	21.767	7.176	8.7683	10.0481
	BT	48.21	44.92	41.81	31.64	29.47	25.53
	R^2	0.9552	0.9822	0.9370	0.9339	0.9134	0.9005
	Δqe	3.92	1.66	5.28	5.45	6.97	8.26

---

3.5.4 Adsorption thermodynamics. Thermodynamic parameters including changes in the standard Gibbs free energy (ΔG), enthalpy (ΔH), entropy (ΔS), were determined by the following equations:³⁵

ΔG = −RT ln K (2)

ΔG = ΔH − TΔS (3)

where R (8.314 J mol⁻¹ K⁻¹) is the gas constant; T (K) is the absolute temperature; K (L mg⁻¹) is the Langmuir constant obtained from the plot of c_e/q_e vs. c_e. The values of ΔH and ΔS can be calculated from the slope and intercept of the plot of ΔG vs. T. From Table 6, the ΔG values for AC and WAC changed from −23.97 to −26.12 kJ mol⁻¹, −16.65 to −18.65 kJ mol⁻¹ with the increasing temperature, respectively, indicating the adsorption of CIP onto AC and WAC were spontaneous and favorable process and they were more spontaneous at higher temperature. The positive value of ΔH confirms an endothermic nature of the adsorption process. The positive value of ΔS suggests the increased randomness at the solid/solution interface.

Table 6 Thermodynamic parameters for the adsorption of CIP on AC and WAC

Sample	T (K)	ΔG (kJ mol^−1)	ΔH (kJ mol^−1)	ΔS (kJ mol^−1 K^−1)
AC	298	−23.97	8.17	0.11
	308	−24.93
	318	−26.12
WAC	298	−16.65	13.27	0.10
	308	−17.59
	318	−18.65

---

4. Conclusion

Two activated carbons have been produced from NMF of WPCB by sodium hydroxide activation in two methods, dry-mixing and wet-mixing. Dry-mixing is the waterless way by direct powder maceration, activated carbon prepared through way is called AC. Wet-mixing is by impregnation within an aqueous solution, activated carbon prepared through way is called WAC. AC had 2304 m² g⁻¹ of surface area and 1.171 cm³ g⁻¹ of total pore volume, while WAC had 1141 m² g⁻¹ of surface area and 0.576 cm³ g⁻¹ of total pore volume. Fourier transform infrared spectra indicated two activated carbons had similar functional groups. For both adsorbents, the adsorption kinetics followed the pseudo-second-order model. The adsorption equilibrium data were very well represented by Langmuir equations. Experiments proved that the method of mixing has great influence on the preparation of activated carbons, and AC showed better characteristics and adsorption properties than WAC.

References

1. P. Hadi, C. Ning, W. Ouyang, M. Xu, C. S. K. Lin and G. McKay, Waste Manag., 2015, 35, 236–246.
2. B. Ghosh, M. K. Ghosh, P. Parhi, P. S. Mukherjee and B. K. Mishra, J. Cleaner Prod., 2015, 94, 5–19.
3. T. Hino, R. Agawa, Y. Moriya, M. Nishida, Y. Tsugita and T. Araki, J. Mater. Cycles Waste Manage., 2009, 11, 42–54.
4. Y. Zhang, S. Liu, H. Xie, X. Zeng and J. Li, Procedia Environ. Sci., 2012, 16, 560–568.
5. I. Birloaga, I. de Michelis, F. Ferella, M. Buzatu and F. Vegliò, Waste Manag., 2013, 33, 935–941.
6. Y. Kan, Q. Yue, J. Kong, B. Gao and Q. Li, Chem. Eng. J., 2015, 260, 541–549.
7. Y.-H. Ke, E.-T. Yang, X. Liu, C.-L. Liu and W.-S. Dong, Carbon, 2013, 60, 563.
8. J. Q. Zhang, S. H. Liu, J. H. Zu, C. Song and C. B. Yu, Appl. Mech. Mater., 2014, 692, 396–400.
9. R.-L. Tseng, J. Colloid Interface Sci., 2006, 303, 494–502.
10. A. C. Johnson, V. Keller, E. Dumont and J. P. Sumpter, Sci. Total Environ., 2015, 511, 747–755.
11. Y. X. Wang, H. H. Ngo and W. S. Guo, Sci. Total Environ., 2015, 533, 32–39.
12. H. Wang, J. Li, C. Ma, Q. Guan, Z. Lu, P. Huo and Y. Yan, Appl. Surf. Sci., 2015, 329, 17–22.
13. P. Wang, Y.-L. He and C.-H. Huang, Water Res., 2010, 44, 5989–5998.
14. M. Brigante and P. C. Schulz, J. Hazard. Mater., 2011, 192, 1597–1608.
15. Y. X. Wang, H. H. Ngo and W. S. Guo, Sci. Total Environ., 2015, 533, 32–39.
16. H. Liu, W. Liu, J. Zhang, C. Zhang, L. Ren and Y. Li, J. Hazard. Mater., 2011, 185, 1528–1535.
17. T. Aysu and M. M. Küçük, Int. J. Environ. Sci. Technol., 2014, 12, 2273–2284.
18. B. H. Hameed, J. Hazard. Mater., 2009, 161, 753–759.
19. Y. Sun, Q. Yue, B. Gao, Q. Li, L. Huang, F. Yao and X. Xu, J. Colloid Interface Sci., 2012, 368, 521–527.
20. A. C. Martins, O. Pezoti, A. L. Cazetta, K. C. Bedin, D. A. S. Yamazaki, G. F. G. Bandoch, T. Asefa, J. V. Visentainer and V. C. Almeida, Chem. Eng. J., 2015, 260, 291–299.
21. M. A. Lillo-Ródenas, J. Juan-Juan, D. Cazorla-Amorós and A. Linares-Solano, Carbon, 2004, 42, 1371–1375.
22. X. Wang, X. Liang, Y. Wang, X. Wang, M. Liu, D. Yin, S. Xia, J. Zhao and Y. Zhang, Desalination, 2011, 278, 231–237.
23. F. Rodriguez-Reinoso, Pure Appl. Chem., 1989, 61, 1859–1866.
24. E. Raymundo-Piñero, P. Azaïs, T. Cacciaguerra, D. Cazorla-Amorós, A. Linares-Solano and F. Béguin, Carbon, 2005, 43, 786–795.
25. M. A. Lillo-Ródenas, D. Cazorla-Amorós and A. Linares-Solano, Carbon, 2003, 41, 267–275.
26. P. Vinke, M. van der Eijk, M. Verbree, A. F. Voskamp and H. van Bekkum, Carbon, 1994, 32, 675–686.
27. M. M. Rao, D. H. Reddy, P. Venkateswarlu and K. Seshaiah, J. Environ. Manage., 2009, 90, 634–643.
28. P. Zhu, Y. Chen, L. Y. Wang and M. Zhou, Waste Manag., 2012, 32, 1914–1918.
29. S. J. Allen, G. McKay and K. Y. H. Khader, Environ. Pollut., 1989, 56, 39–50.
30. B. H. Hameed, I. A. W. Tan and A. L. Ahmad, Chem. Eng. J., 2008, 144, 235–244.
31. S. A. C. Carabineiro, T. Thavorn-Amornsri, M. F. R. Pereira and J. L. Figueiredo, Water Res., 2011, 45, 4583–4591.
32. Q. Wu, Z. Li, H. Hong, K. Yin and L. Tie, Appl. Clay Sci., 2010, 50, 204–211.
33. S. A. C. Carabineiro, T. Thavorn-amornsri, M. F. R. Pereira, P. Serp and J. L. Figueiredo, Catal. Today, 2012, 186, 29–34.
34. Q. Wu, Z. Li, H. Hong, R. Li and W.-T. Jiang, Water Res., 2013, 47, 259–268.
35. W. Liu, J. Zhang, C. Zhang, Y. Wang and Y. Li, Chem. Eng. J., 2010, 162, 677–684.

---