Mussel inspired functionalization of carbon nanotubes for heavy metal ion removal

Abstract

Carbon nanotubes (CNTs) have been widely used as adsorbents to remove various environmental pollutants because of their unique one dimensional structure, large surface areas and amount of micropores. However, the adsorption capacity of unmodified CNTs toward heavy metal ions is still limited due to their poor dispersibility and lack of functional groups. In this work, a novel strategy has been developed to prepare polyethylenimine functionalized CNTs via the combination of mussel inspired chemistry and the Michael addition reaction. The successful preparation of CNTs with amine groups was confirmed by a series of characterization measurements such as transmission electron microscopy, Fourier transform infrared spectroscopy, and thermal gravimetric analysis. Furthermore, the adsorption application of these amine functionalized CNTs toward Cu²⁺ was also examined. The effects of various factors including contact time, pH values, temperature and initial Cu²⁺ concentrations on the adsorption capability of the amine functionalized CNTs were also investigated. Langmuir and Freundlich models were used for thermomechanical analysis. The pseudo-first-order, pseudo-second-order and intra-particle diffusion models were used for the kinetics analysis. The results demonstrated that CNTs can be successfully functionalized with amine groups through a rather facile and mild bioinspired strategy. These amine functionalized CNTs exhibited a much enhanced adsorption efficiency toward Cu²⁺. Given the strong and versatile adhesion of PDA to various materials, the bioinspired strategy described in this work could also be utilized for the fabrication of many other nanocomposites for environmental applications.

---

1. Introduction

Under the current circumstances, human beings pay great attention to environmental contamination, certainly including water pollution.¹ Heavy metal ions, which exist in various industrial wastewaters, are attracting growing attention due to their high toxicity, non-biocompatibility and easy accumulation in living organisms.^2–7 Among them, the copper ion (labelled as Cu²⁺), is considered to be one of the most poisonous ions, normally dissolved into wastewater and then released into the environment. Although it is a significant trace nutrient for the health of the human body, when consumed in a large dose for a long period it may lead to vomiting, cramps, skin rash or even death.^8,9 As we all know, large amounts of Cu²⁺ are produced in the fields of smelting, metal processing, machinery manufacturing, organic synthesis and other industrial wastewater.^10,11 Among them, the metal processing and electroplating factories discharge waste water containing the highest copper content, which has caused serious impacts on living organisms. Therefore removal of Cu²⁺ from wastewater is necessary and urgent.

Due to the hazardous consequences of high concentrations of Cu²⁺ for human life, in order to resolve these problems, many high-efficiency methods for the removal of Cu²⁺ from wastewater have been reported, including ion exchange, chemical precipitation, ultrafiltration, reverse osmosis, adsorption, electrodialysis, etc.^12–17 Among these techniques, the typical strategy of adsorption is very prominent and draws much attention for its high efficiency, simple operation and low cost. Additionally, with the development of nanotechnology, a large number of nanomaterials, especially carbonaceous nanomaterials (e.g. carbon nanotubes (CNTs) and graphene oxide), are used as adsorbents because of their large pore volume, large specific surface area and stable covalent chemical bonds.^18,19 CNTs, a class of popular carbon nanomaterials, have been widely applied in various fields such as drug delivery, machine manufacturing, environmental protection and others because of their great mechanical properties, high chemical stability, unique electronic properties and diameter ratio.²⁰ Recently, accompanying the in-depth research of CNTs, the excellent performances of the huge specific surface areas and loose structure of CNTs have attracted much attention in the adsorption field, in which CNTs could serve as high-efficiency agents to remove other heavy ions.^21–31 Although they possess these perfect performances, unfortunately, poor dispersibility and lack of functional groups inevitably restricted the adsorption capacity of CNTs toward heavy metal ions. In recent years, many strategies for the surface modification of CNTs to enhance their dispersibility in aqueous solution have been developed.^32–39 However, the commonly adopted noncovalent and covalent strategies still have some drawbacks, which include the instability of noncovalent methods, and the low efficiency, complexity and time consumption of the covalent strategy. Therefore, the development of facile surface modification strategies for the preparation of CNTs with good adsorption performance is still highly desirable.

As a versatile surface modification strategy, mussel-inspired chemistry has attracted increasing attention for wide applications ranging from biological imaging and cancer treatment to energy conversion and environmental protection.^40–57 Since this novel strategy was discovered by Lee in 2007, mussel-inspired chemistry has become a versatile and facile surface modification method for any material, including metal and non-metal materials, regardless of their composition, shape and size, because of the strong and universal adhesion of polydopamine (PDA), which is formed via the self-oxidation of dopamine in alkaline solution. Furthermore, after coating PDA coating onto the material surface, many reactive sites are introduced onto the surface of material, which can further be covalently conjugated with amino- or thiol-containing molecules.⁴³ For example, our previous work reported the surface modification of CNTs with thiol-containing small organic molecules and amino-terminated polymers to increase the dispersibility of the CNT materials via a combination of mussel-inspired chemistry and the Michael addition reaction. These functional CNTs showed obviously enhanced dispersion in aqueous and some organic solutions.^58–61 Polyethylenimine (PEI), as a hydrophilic organic molecule with abundant amino groups, has been widely used in the surface modification of adsorbents due to its numerous amine groups, including primary, secondary and tertiary amines, which exhibit perfect adsorption ability for heavy metals.⁶² Therefore, inspired by these facile surface functional methods, amine functionalized CNTs can be prepared through the Michael addition reaction between PDA and PEI.

In this work, the stiff CNT surfaces were perfectly modified with hydrophilic PEI to endow their excellent dispersibility in water and adsorption ability toward Cu²⁺ via the combination of mussel-inspired chemistry and the Michael addition reaction. The specific procedure is shown in Scheme 1: two mild steps were involved in the process for the preparation of the functionalized adsorbents. First, the PDA coating was formed on the surface of the CNTs in aqueous solution (pH = 8.5) via mussel inspired chemistry. Afterwards, PEI molecules were efficiently conjugated onto the surface of the CNTs functionalized with PDA through the Michael addition reaction. Finally, the obtained amine functionalized CNTs were applied to remove Cu²⁺. The factors influencing the removal of Cu²⁺, including pH, contact time, temperature and initial Cu²⁺ concentration, were investigated in this work.

Scheme 1 Schematic representation of the preparation of CNT-PDA-PEI via the combination of mussel-inspired chemistry and the Michael addition reaction.

2. Experiment

2.1 Materials

CNTs from the Nanopowder Co. were synthesized by chemical vapor deposition. The diameter of the CNTs is about 30–50 nm according to the manufacturer. The dopamine hydrochloride was purchased from Sangon Co. Tris-(hydroxymethyl)-aminomethane (Tris) (>99%) was obtained from Tianjin Heowns Biochem. LLC. Co. PEI (molecular weight 800 Da) and copper nitrate trihydrate (>99.99%) were purchased from Aladdin Industrial Co. Deionized water was prepared for usage in solutions.

2.2 Characterization

The morphology of the modified CNT-PDA-PEI adsorbent was measured by transmission electron microscopy (TEM). The TEM images were obtained using a Hitachi 7650B microscope operated at 80 kV by using a little glob of nanoparticle ethanol suspension on a carbon-coated copper grid as the TEM specimen. Fourier transform infrared spectroscopy (FT-IR) was used to confirm the structure, composition and chemical functional groups of the modified CNT-PDA-PEI. The Fourier transform infrared (FT-IR) spectra were obtained using a Nicolet 5700 (Thermo Nicolet corporation). Thermal gravimetric analysis (TGA), which was conducted on a TA instrument Q50 with a heating rate of 20 °C min⁻¹, was used for determination of the thermal properties of the modified CNT-PDA-PEI under an air atmosphere.

2.3 Synthesis of CNT-PDA

The synthesis method of CNT-PDA nanocomposites was as follows in detail: 1 g of CNTs and 1 g of dopamine hydrochloride were simultaneously added into a 500 mL flask with Tris buffer solution (pH = 8.5, 10 mM) into the flask to make a 250 mL solution. After that, the suspensions were treated with a ultrasonicator for about 5 min and magnetically stirred for 8 h. Finally they were centrifuged at 8000 rpm for 10 min, washed 3 times with deionized water and dried at 323 K for 24 h.

2.4 CNT-PDA-PEI

CNT-PDA-PEI was synthesized by the same method mentioned above. 1 g of PEI and 1 g of CNT-PDA were added into a 500 mL flask using deionized water with Tris buffer solution, then ultrasonicated for 5 min and stirred for 8 h, centrifuged, washed with pure water three times and dried under vacuum for further adsorption experiments.

2.5 Adsorption capacity of CNT-PDA-PEI for Cu²⁺

Batch adsorption experiments were used to study the removal of Cu²⁺ under the influence of the following parameters: contact temperature, pH value, contact time and the initial Cu²⁺ concentration. In general, 10 mg adsorbent was added into 50 mL of the Cu²⁺ solution (10 mg L⁻¹) in the 50 mL sample tube. The equilibrium Cu²⁺ concentrations were determined using the ultraviolet and visible spectrophotometer. The effect of contact temperature was studied at different temperatures (303, 313, 323, 333, 343, 353 K) with the same amount of the adsorbent and the same Cu²⁺ concentration. The effect of pH was studied by changing the Cu²⁺ solution pH from 2 to 10 by adding 0.1 M HCl and 0.1 M NaOH solutions. The influence of Cu²⁺ concentration was tested using different initial Cu²⁺ concentrations in the range of 10 mg L⁻¹ to 120 mg L⁻¹. The final adsorption capacity (Q_e, mg g⁻¹) and the Cu²⁺ removal efficiency (R%) were analyzed by the following equations, respectively:

where V (mL) is the Cu²⁺ solution volume, m (mg) is the adsorbent dose, C₀ (mg L⁻¹) is the Cu²⁺ initial concentration and C_t is the equilibrium Cu²⁺ concentration. Q_e (mg g⁻¹) is the adsorption capacity at the equilibrium time, and R (%) is the removal efficiency of the Cu²⁺ solution.

3. Results and discussion

3.1 Characterization of the CNT nanocomposites

The surface modified CNTs with the organic molecule abundant in amino groups (PEI) were prepared via a novel strategy of mussel-inspired chemistry and the Michael addition reaction. The TEM images of the CNT samples demonstrated that the PDA coating and PEI perfectly covered the surface of CNTs. As we can see from Fig. 1A, the diameter of the pristine CNTs was about 30 to 50 nm, which was consistent with information provided from the manufacturers. The surfaces of the pristine CNTs are especially smooth, from the observation of Fig. 1A. After modifying PDA onto the CNT surface via mussel-inspired chemistry, as shown in Fig. S1,† the CNT surface changed to slightly coarse from the smooth surface, which provided direct evidence of the successful modification of the CNT surface with PDA coating as well. These results successfully demonstrated that the strong adhesion of PDA is efficient for inorganic materials. The phenomenon of Fig. S1† could illustrate that mussel-inspired chemistry is suitable to any material. After PEI molecules were covalently conjugated onto the CNT-PDA surface via a typical Michael addition reaction, a blurry coating on the CNT surface can be observed (as marked with red arrows in Fig. 1B and S2†). Based on the enlarged TEM image of CNT-PDA-PEI, the thickness of the PDA coating is about 8 ± 1 nm. These results provided evidence that PEI was grafted onto the surface of the CNTs due to the role of the Michael addition reaction. From the TEM analysis, we concluded the successful modification of CNTs with hydrophilic PEI to enhance their dispersion via the combination of mussel inspired chemistry and the Michael addition reaction. More importantly, the novel strategy described in this work is also suitable other materials to endow them with excellent properties.

Fig. 1 Representative TEM images of CNT samples. (A) Pristine CNTs, (B) CNT-PDA-PEI. These primarily demonstrate the successful modification of CNTs with PDA and PEI via the combination of mussel-inspired chemistry and the Michael addition reaction.

In addition to the high-resolution TEM characterization technique, the typical FT-IR technique was also employed to determine whether the surfaces of the CNTs were successfully modified using PDA and PEI. As shown in Fig. 2, although there were no particular peaks that could be observed for the pristine CNTs, after PDA was coated onto the CNT surface because of its strong adhesion to any material, the characteristic peak located at 3450 cm⁻¹ was observed, which was attributed to the stretching vibration of –NH₂ bonds. Additionally, the characteristic peak that appeared at 2850 cm⁻¹ could be ascribed to the stretching vibration of the –CH₂ existing in the catechol. These results demonstrated that a PDA coating is definitely formed on the CNT surface by the facile mussel-inspired chemistry surface modification method. Based on the typical Michael addition reaction, the PEI molecules could be facilely conjugated onto the surface of CNT-PDA; the powerful evidence which could be used to verify the successful conjugation of PEI to the CNT-PDA surface is that a new peak appeared at 2930 cm⁻¹, from –CH₃ bonds. Furthermore, the two characteristic peaks respectively located at 1630 and 1410 cm⁻¹ could be ascribed to the vibration of C–N (simple harmonic vibration) and C–C (flexible vibration) bonds. Therefore, the FT-IR spectra of the CNT samples demonstrated that functionalized CNTs were prepared using PEI based on mussel-inspired chemistry and the Michael addition reaction.

Fig. 2 FT-IR spectra for CNTs, CNT-PDA, and CNT-PDA-PEI nanotubes.

The relative content of PDA and PEI conjugated onto the surface of the CNTs could be confirmed by TGA measurements. As we can see from the TGA curves of the CNT samples (Fig. S3†), the weight loss of the pristine CNTs was approximately 1.17% when the temperature ascended from 100 to 500 °C under nitrogen atmosphere. The slight weight loss of the pristine CNTs at high temperature suggested that the CNTs possess excellent thermal stability. On the other hand, a pronounced weight loss of the CNT samples (weight loss: 19.30%) can also be observed in Fig. S3† when the temperature is raised from 500 to 600 °C, suggesting that the decomposition temperature of CNTs is 500 °C. However, after the formation of PDA on the CNT surface due to the self-polymerization of DA in the alkaline solution, the weight loss rose to 9.81% when the temperature was increased from 100 to 500 °C. Compared to the pristine CNTs, the change in weight loss of CNT-PDA at the same temperature could be explained as the oxygen content being enhanced because of the introduction of the PDA coating on the CNT surface, which evidenced the successful formation of the PDA coating on the CNT surface by mussel-inspired chemistry. The content of the PDA coating on the CNT surface could be calculated as 8.64%. After adding the PEI to the surface of CNT-PDA via the Michael addition reaction, the weight loss of CNT-PDA-PEI samples was increased to 40.1% in an air atmosphere (Fig. S3†). Compared with CNT-PDA, the weight loss of CNT-PDA-PEI was significantly increased from 19.3 to 40.1%, further demonstrating that PEI was successfully conjugated onto the CNT-PDA surface via a typical Michael addition reaction.

3.2 Adsorption experiment studies

3.2.1 Effect of contact time and adsorption kinetics. The effect of contact time on Cu²⁺ removal by the pristine CNTs and CNT-PDA-PEI is shown in Fig. 3. Obviously, the modified CNT-PDA-PEI nanocomposites show better adsorption performance than the pristine CNTs, which demonstrated that the PEI molecules were covalently conjugated onto the surface of CNT-PDA via a typical Michael addition reaction. The Cu²⁺ equilibrium adsorption capacity of CNTs is just 25.4 mg g⁻¹ while the Q_t of the CNT-PDA-PEI reaches up to 34.6 mg g⁻¹, greater by a large margin, which is due to the abundant amine groups existing in the PEI macromolecules on the surface of the CNTs. Therefore the dramatic improvement of the adsorption capacity also suggested the perfect adsorption property of CNTs modified with PEI for Cu²⁺ removal. More importantly, the obvious enhancement of the adsorption of CNT-PDA-PEI versus the CNT material provides direct evidence of the feasibility of the novel strategy described in this work for the surface modification of materials. The above results suggested that the amine groups on the adsorbent (CNT-PDA-PEI) play an important role for the enhancement of adsorption capacity, indicating that a novel adsorption mechanism was involved in the adsorption procedure of CNT-PDA-PEI. It is well known that the introduction of functional groups such as carboxyl groups and amino groups on CNTs can improve the adsorption capability of these adsorbents through increasing the active adsorption sites.⁶³ This enhancement adsorption of is mainly due to the coordination of functional groups with cations. Most of these interactions are not specific interactions. Therefore, the CNT-PDA-PEI are not selective adsorbents toward Cu²⁺.

Fig. 3 The effect of contact time on Cu²⁺ adsorption with an initial Cu²⁺ concentration of 10 mg L⁻¹ at room temperature.

The non-linear form of the pseudo-first-order kinetic model is given as follows:

Q_t = Q_e(1 − e^−k₁t)

where Q_t (mg g⁻¹) is the quantity of Cu²⁺ adsorbed by CNT-PDA-PEI at different contact times t (min), Q_e (mg g⁻¹) is the equilibrium adsorption capacity, and k₁ (min⁻¹) is the pseudo-first-order equation rate constant. The values of these parameters and k₁ and the correlation coefficient (R²) are listed in Table 1. The following equations are often used to describe the non-linear form of the pseudo-second-order kinetic model.

h = k₂Q_e²

where Q_e (mg g⁻¹) and Q_t (mg g⁻¹) are the adsorption capacity of the modified adsorbent CNT-PDA-PEI at equilibrium and different times t (min), respectively. k₂ (g mg⁻¹ min⁻¹) is the rate constant of the pseudo-second-order model, and h (mg g⁻¹ min) is the initial adsorption rate of the Cu²⁺ removal process. All the parameters and R² are also listed in Table 1. This study used intra-particle diffusion model to further test the reaction rate of the adsorption process, which can be described by the equation as follows:

Q_t = k_pt^0.5

where Q_t (mg g⁻¹) is the adsorption capacity at any time t (min), as described above, k_p (mg g⁻¹ min^−1/2) is the rate constant of the intra-particle diffusion model. The value of k_p (mg g⁻¹ min^−1/2) and R² are listed in Table 1. According to the results summarized in Table 1, the correlation coefficient (R²) from the pseudo-first-order model is 0.994, which is much higher than the correlation coefficient (R² = 0.841) from the pseudo-second-order model. These results indicated that the adsorption kinetics followed the pseudo-first-order kinetic model better (Fig. 4). However, intra-particle diffusion is the rate-controlling step of the adsorption process due to the relatively high correlation coefficient (R² = 0.985) from the intra-particle diffusion kinetic model. In summary, the several steps which happened in the adsorption process could be classified as follows: the instantaneous step occurred at first on the apparent surface, and the gradual adsorption step happened as the rate-controlling reaction process. The last equilibrium step with low adsorption efficiency could be attributed to the decrease of the monolayer site and the reduction of the Cu²⁺ concentration. As is known to all, the hydrolysis capacity of the Cu²⁺ ion and the competitive adsorption of coexisting matter in aqueous solution both are influencing factors of the Cu²⁺ removal capacity on the new composite adsorbent CNT-PDA-PEI. Furthermore, the chemical and physical properties of the chosen adsorbent also have an impact on the adsorption process. Incidentally the high adsorption rate in this study may be owing to chemisorption. Therefore, the complexation chemical reaction is expected in the adsorption.

Table 1 Adsorption kinetics data of pseudo-first-order, pseudo-second-order and intra-particle diffusion model for Cu²⁺ adsorption on CNT-PDA-PEI. Initial Cu²⁺ concentration: 10 mg L⁻¹

Model	Parameter	Value
Pseudo-first-order	Qe(cal) (mg g^−1)	34.1
	k2 (min^−1)	0.0473
	R^2	0.994
Pseudo-second-order	Qe(cal) (mg g^−1)	79.1
	k2 (g mg^−1 min g^−1)	0.0000799
	h (mg g^−1 min g^−1)	0.500
	R^2	0.841
Intra-particle diffusion	kp (mg g^−1 min^−1/2)	549 797
	R^2	0.985

---

Fig. 4 Different kinds of adsorption kinetics of CNT-PDA-PEI. Values of the data fitted by pseudo-first-order, pseudo-second-order and intra-particle diffusion models. The experiment proceeded at room temperature, pH = 7.0 and 10 mg L⁻¹ initial Cu²⁺ concentration.

3.2.2 Effect of temperature and thermodynamic study. Temperature is one of indispensable factors in adsorption experiments. The effect of temperature on the removal of Cu²⁺ using CNT-PDA-PEI was tested at different temperatures ranging from 303 to 353 K with the other conditions unchanged. The thermodynamic parameters including enthalpy change (ΔH⁰), Gibbs free energy (ΔG⁰) and entropy change (ΔS⁰) were determined by using the following equations:

ΔG⁰ = −RT ln Kα

where Kα is the thermodynamic equilibrium constant, R represents the gas constant (8.314 J mol⁻¹ K), and T is the solution temperature (K). ΔH⁰ and ΔS⁰ were calculated from the slope and intercept of the Van’t Hoff plots of ln Kα with 1/T, respectively. The obtained thermodynamic parameters are recorded in Table 2. As shown in Fig. 5, the Cu²⁺ adsorption capacity increased clearly from 21.6 to 41.0 mg g⁻¹ with the increase of temperature from 303 to 353 K. In addition, as can be seen from the thermodynamic parameters in Table 2 calculated from the above, the positive values of ΔH⁰ and the increasing trend of the adsorption capacity with the temperature increase both signified that adsorption process using CNT-PDA-PEI is an endothermic process. The negative values of ΔG⁰ signified the spontaneous nature of the process of Cu²⁺ adsorption on CNT-PDA-PEI.

Table 2 Data of the thermodynamic parameters for the adsorption of Cu²⁺ onto the CNTs-PDA-PEI nanocomposites

T (K)	ΔG^0 (kJ mol^−1 K^−1)	ΔH^0 (kJ mol^−1 K^−1)	ΔS^0 (kJ mol^−1 K^−1)
303	−3.38	34.4	0.113
313	−4.32	35.5
323	−5.85	36.6
333	−6.70	37.8
343	−7.60	38.9
353	−9.18	40.0

---

Fig. 5 (A) The effect of temperature on Cu²⁺ removal using CNT-PDA-PEI. (B) Van’t Hoff plot for the Cu²⁺ adsorption process.

3.2.3 Effect of pH. The pH value is of great influence, reflecting on both the conversion of the metal ions from aqueous solution and the protonation degree of the amine groups on CNT-PDA-PEI. In order to test the effect of the initial pH of the Cu²⁺ solution on the adsorption of CNT-PDA-PEI, in this work different initial pH solutions in the range of 3 to 12 at room temperature were made. As shown in Fig. 6, the adsorption capacity was increased from 4.50 to 47.4 mg g⁻¹ as the pH was increased from 3 to 10, and then remained approximately stable at pH > 10. For the first rising stage, adsorption capacity was low due to the protonation of the amine groups on the CNT-PDA-PEI in acidic solutions; however, the adsorption capacity increased because the amine group deprotonation was weaken with increasing pH values. Gradually in an alkaline environment, on account of the porous structure and abundant deprotonated –NH₂ groups on the surface of CNT-PDA-PEI, the adsorption capacity was considerably greater, as well as the coordinating ability between the adsorbent and Cu²⁺. On the other hand, however, part of the PDA structure can be damaged in a very alkaline environment especially when pH > 11, which may hinder the improvement of the adsorption capacity and cause it to remain approximately the same. As for the removal efficiency in Fig. 6, there was an obvious enhancement from 9.09 to 95.7%, which further indicated that the modified CNT-PDA-PEI adsorbent shows a great Cu²⁺ removal ability.

Fig. 6 The effect of pH on Cu²⁺ removal using CNT-PDA-PEI with 10 mg L⁻¹ initial Cu²⁺ concentration at room temperature.

3.2.4 Adsorption isotherms. The adsorption isotherms manifested the way that the adsorbents interacted with heavy metal ions or other adsorbates, and also described the adsorption capacity, adsorption strength and the adsorption status. In this work, two significant adsorption isotherm models, namely the Langmuir and Freundlich isotherms, were used to analyze the results of the equilibrium isotherms of Cu²⁺ solution using CNT-PDA-PEI as the adsorbent at different initial Cu²⁺ concentrations in a neutral environment and room temperature (Fig. 7).

Fig. 7 Adsorption isotherms of Cu²⁺ on CNT-PDA-PEI. Experimental conditions: adsorbent dose 10 mg, initial Cu²⁺ concentrations from 10 mg L⁻¹ to 100 mg L⁻¹. The values of the data fitted by the Langmuir model and Freundlich model.

The Langmuir isotherm is also called the adsorption model of monomolecular layer, and just as the name implies the isotherm is based on the assumption that each adsorption site is distributed homogeneously on the surface of solid materials and can only adsorb a single molecule. The adsorption model can be expressed by the following equation:

where Q_m (mg g⁻¹) is the maximum amount of adsorbate during the whole adsorption process, Q_e (mg g⁻¹) is the adsorption capacity at adsorption equilibrium, C_e (mg L⁻¹) is the equilibrium concentration of the adsorbed Cu²⁺ solution, and b (l mg⁻¹) is the adsorption index. The higher the b value, the stronger the adsorption process.

The Freundlich isotherm model is an empirical isothermal equation based on experimental results which is used to describe heterogeneous systems. The form of the Freundlich equation can be represented as follows:

where Q_e (mg g⁻¹) is the equilibrium adsorption capacity, C_e (mg L⁻¹) is the equilibrium concentration of the Cu²⁺ solution , K_F [(mg g⁻¹) (l mg⁻¹)^1/n] is an correlation index reflecting the adsorption capacity, and n is the adsorption constant as well as indicating the adsorption intensity. The above adsorption isotherm parameters and the correlation coefficients (R²) of the Langmuir and Freundlich models are listed in Table 3. In this work, the values of Q_m and b from the Langmuir model for CNT-PDA-PEI are 70.9 mg g⁻¹ and 0.0511 L mg⁻¹, respectively, and R² from the Langmuir model is 0.991. As for the Freundlich model, the value of K_F is 12.2 (mg g⁻¹) (L mg⁻¹) ^1/n, n is 2.84, and R² is 0.89 at 298 K.

Table 3 Adsorption isotherm data of the Langmuir and Freundlich models for Cu²⁺ adsorption on CNT-PDA-PEI. Temperature: 298 K

Isotherm	Parameter	Value
Langmuir	Q0 (mg g^−1)	70.9
	b (L mg^−1)	0.0511
	R^2	0.991
Freundlich	KF [(mg g^−1) (L mg^−1)^1/n]	12.2
	n	2.84
	R^2	0.893

---

As the results in Table 3 show, the Langmuir model has a higher value of correlation coefficient (R²) than the Freundlich model, which represents the Langmuir isotherm model being more aligned with the experimental data then the Freundlich model. Additionally, the value of n (2.84) from the Freundlich isotherm model means that 1/n is approximately equal to 0.35, between the numerical values of 0.1 to 0.5, which illustrates the favorable adsorption of the Cu²⁺ solution on CNT-PDA-PEI. According to all of the above experimental data and computed results, we can consequently conclude that the active adsorption sites are distributed homogeneously on the modified CNT-PDA-PEI nanocomposites, and the Cu²⁺ removal process complies with the monolayer adsorption. From the Langmuir model, the maximum amount of the adsorption capacity by modified CNT-PDA-PEI reached about 70.9 mg g⁻¹. Compared with the pristine CNTs and CNT-PDA, the results show the better adsorption efficiency of CNT-PDA-PEI for the removal of Cu²⁺ from aqueous solution.

3.2.5 Effect of adsorbent dosage. As shown in Fig. 8, increasing the adsorbent dosage resulted in an increase in the percentage removal of Cu²⁺. Afterwards, with the continuous increase of the CNT-PDA-PEI, the removal percentage R (%) reached up to about 92%, which was close to a hundred percent, and remained steady. The adsorption capacity variation could be connected to the adsorbent concentration change, which may due to the variation of the modified adsorbent surface area and functional groups. Some methods for the practical adsorption applications of these CNT based adsorbents may be adopted. For example, we can fill these modified CNT based adsorbents into a column and the wastewater can be purified by the adsorbent filled column. Additionally, these adsorbents can be conjugated with magnetic nanoparticles, and therefore these hybrid adsorbents can separated from water using magnetic field.

Fig. 8 Effect of adsorbent dosage on removal of Cu²⁺. Experimental conditions: reaction time: 120 min, Cu²⁺ concentration: 10 mg L⁻¹, room temperature and neutral pH.

4. Conclusion

In summary, a facile, efficient and rapid approach was reported for the preparation of an excellent absorbent with great dispersibility based on attractive CNTs for the removal of Cu²⁺ from aqueous solution in this work via the combination of mussel-inspired chemistry and the Michael addition reaction method. Characterization techniques including TEM, FT-IR and TGA were employed to ensure the successful surface modification of the CNTs with PDA and PEI. With regard to the adsorption experiments, the effects of contact time, pH, temperature and initial concentrations were investigated. Additionally, according to the adsorption experiments for Cu²⁺ removal at different conditions (temperature, time, pH, Cu²⁺ concentration), the adsorption capacity of CNT-PDA-PEI is increased with increasing time, pH and temperature, suggesting the improved adsorption capability of CNTs modified with PEI. Furthermore, the positive value of ΔH° and the negative value of ΔG° indicated that the removal of Cu²⁺ using CNT-PDA-PEI as an adsorbent is endothermic and spontaneous. The parameters for the Langmuir and Freundlich isotherms were also calculated. The authentic purpose of this communication is to aim at providing a facile and versatile method for the fabrication of high efficiency adsorbents, which should be of great importance for environmental management.

Acknowledgements

This research was supported by the National Science Foundation of China (No. 21134004, 21201108, 51363016, 21474057), and the National 973 Project (No. 2011CB935700).

References

1. V. K. Gupta, I. Ali, T. A. Saleh, A. Nayak and S. Agarwal, RSC Adv., 2014, 2, 6380–6388.
2. A. Bhatnagar, F. Kaczala, W. Hogland, M. Marques, C. A. Paraskeva, V. G. Papadakis and M. Sillanpää, Environ. Sci. Pollut. Res., 2014, 21, 268–298.
3. F. Fu and Q. Wang, J. Environ. Manage., 2011, 92, 407–418.
4. K. Rana, M. Shah and N. Limbachiya, Int. J. Adv. Res. Comput. Sci. Software Eng., 2014, 1, 55–59.
5. L. Ranjbar, Y. Yamini, A. Saleh, S. Seidi and M. Faraji, Microchim. Acta, 2012, 177, 119–127.
6. S. K. Das, I. Shome and A. K. Guha, RSC Adv., 2012, 2, 3000–3007.
7. K. Yang, W. Wei, L. Qi, W. Wu, Q. Jing and D. Lin, RSC Adv., 2014, 4, 46122–46125.
8. Y. Liu, M. Chen and H. Yongmei, Chem. Eng. J., 2013, 218, 46–54.
9. B. Murphy and B. Hathaway, Coord. Chem. Rev., 2003, 243, 237–262.
10. M. Kazemipour, M. Ansari, S. Tajrobehkar, M. Majdzadeh and H. R. Kermani, J. Hazard. Mater., 2008, 150, 322–327.
11. H. A. Qdais and H. Moussa, Desalination, 2004, 164, 105–110.
12. A. Da̧browski, Z. Hubicki, P. Podkoscielny and E. Robens, Chemosphere, 2004, 56, 91–106.
13. Y.-P. Zhu, Y.-L. Liu, T.-Z. Ren and Z.-Y. Yuan, RSC Adv., 2014, 4, 16018–16021.
14. S. Zhang, J. Li, T. Wen, J. Xu and X. Wang, RSC Adv., 2013, 3, 2754–2764.
15. S. Dib, M. Boufatit, S. Chelouaou, F. Sadi-Hassaine, J. Croissant, J. Long, L. Raehm, C. Charnay and J.-O. Durand, RSC Adv., 2014, 4, 24838–24841.
16. S. Yang, D. Shao, X. Wang and M. Nagatsu, RSC Adv., 2014, 4, 4856–4863.
17. B. Zou, K. Chen, Y. Wang, C. Niu and S. Zhou, RSC Adv., 2015, 5, 22973–22979.
18. S.-T. Yang, Y. Chang, H. Wang, G. Liu, S. Chen, Y. Wang, Y. Liu and A. Cao, J. Colloid Interface Sci., 2010, 351, 122–127.
19. C. Cheng, S. Li, J. Zhao, X. Li, Z. Liu, L. Ma, X. Zhang, S. Sun and C. Zhao, Chem. Eng. J., 2013, 228, 468–481.
20. R. H. Baughman, A. A. Zakhidov and W. A. de Heer, Science, 2002, 297, 787–792.
21. J. L. Gong, B. Wang, G. M. Zeng, C. P. Yang, C. G. Niu, Q. Y. Niu, W. J. Zhou and Y. Liang, J. Hazard. Mater., 2009, 164, 1517–1522.
22. S. A. Kosa, G. Al Zhrani and M. A. Salam, Chem. Eng. J., 2012, 181, 159–168.
23. Y. Li, F. Liu, B. Xia, Q. Du, P. Zhang, D. Wang, Z. Wang and Y. Xia, J. Hazard. Mater., 2010, 177, 876–880.
24. Y. H. Li, J. Ding, Z. Luan, Z. Di, Y. Zhu, C. Xu, D. Wu and B. Wei, Carbon, 2003, 41, 2787–2792.
25. G. P. Rao, C. Lu and F. Su, Sep. Purif. Technol., 2007, 58, 224–231.
26. S. Rosenzweig, G. A. Sorial, E. Sahle-Demessie, D. C. McAvoy and A. A. Hassan, Water, Air, Soil Pollut., 2014, 225, 1–17.
27. E. Salehi, S. Madaeni, L. Rajabi, V. Vatanpour, A. Derakhshan, S. Zinadini, S. Ghorabi and H. A. Monfared, Sep. Purif. Technol., 2012, 89, 309–319.
28. A. Stafiej and K. Pyrzynska, Sep. Purif. Technol., 2007, 58, 49–52.
29. C. H. Wu, J. Colloid Interface Sci., 2007, 311, 338–346.
30. C. Zhang, J. Sui, J. Li, Y. Tang and W. Cai, Chem. Eng. J., 2012, 210, 45–52.
31. G. Zhao, T. Wen, C. Chen and X. Wang, RSC Adv., 2012, 2, 9286–9303.
32. F. Gojny, M. Wichmann, U. Köpke, B. Fiedler and K. Schulte, Compos. Sci. Technol., 2004, 64, 2363–2371.
33. S. W. Kim, T. Kim, Y. S. Kim, H. S. Choi, H. J. Lim, S. J. Yang and C. R. Park, Carbon, 2012, 50, 3–33.
34. P. Liu, Eur. Polym. J., 2005, 41, 2693–2703.
35. M. Gkikas, B. P. Das, M. Tsianou, H. Iatrou and G. Sakellariou, Eur. Polym. J., 2013, 49, 3095–3103.
36. X. Zhang, K. Wang, M. Liu, X. Zhang, L. Tao, Y. Chen and Y. Wei, Nanoscale, 2015, 7, 11486–11508.
37. Q. Wan, M. Liu, J. Tian, F. Deng, G. Zeng, Z. Li, K. Wang, Q. Zhang, X. Zhang and Y. Wei, Polym. Chem., 2015, 6, 1786–1792.
38. B. Yang, Y. Zhao, X. Ren, X. Zhang, C. Fu, Y. Zhang, Y. Wei and L. Tao, Polym. Chem., 2015, 6, 509–513.
39. X. Ren, B. Yang, Y. Zhao, X. Zhang, X. Wang, Y. Wei and L. Tao, Polymer, 2015, 64, 210–215.
40. Y. Cao, X. Zhang, L. Tao, K. Li, Z. Xue, L. Feng and Y. Wei, ACS Appl. Mater. Interfaces, 2013, 5, 4438–4442.
41. J. Cui, Y. Yan, G. K. Such, K. Liang, C. J. Ochs, A. Postma and F. Caruso, Biomacromolecules, 2012, 13, 2225–2228.
42. S. M. Kang, S. Park, D. Kim, S. Y. Park, R. S. Ruoff and H. Lee, Adv. Funct. Mater., 2011, 21, 108–112.
43. H. Lee, S. M. Dellatore, W. M. Miller and P. B. Messersmith, Science, 2007, 318, 426–430.
44. J. Wu, L. Zhang, Y. Wang, Y. Long, H. Gao, X. Zhang, N. Zhao, Y. Cai and J. Xu, Langmuir, 2011, 27, 13684–13691.
45. S. H. Yang, S. M. Kang, K. B. Lee, T. D. Chung, H. Lee and I. S. Choi, J. Am. Chem. Soc., 2011, 133, 2795–2797.
46. Q. Zhu and Q. Pan, ACS Nano, 2014, 8, 1402–1409.
47. H. Gao, Y. Sun, J. Zhou, R. Xu and H. Duan, ACS Appl. Mater. Interfaces, 2013, 5, 425–432.
48. Y. Liu, K. Ai and L. Lu, Chem. Rev., 2014, 114, 5057–5115.
49. Q. Ye, F. Zhou and W. Liu, Chem. Soc. Rev., 2011, 40, 4244–4258.
50. C. Cheng, S. Nie, S. Li, H. Peng, H. Yang, L. Ma, S. Sun and C. Zhao, J. Mater. Chem. B, 2013, 1, 265–275.
51. C. Cheng, S. Sun and C. Zhao, J. Mater. Chem. B, 2014, 2, 7649–7672.
52. H. Hu, B. Yu, Q. Ye, Y. Gu and F. Zhou, Carbon, 2010, 48, 2347–2353.
53. Q. Wan, J. Tian, M. Liu, G. Zeng, Q. Huang, K. Wang, Q. Zhang, F. Deng, X. Zhang and Y. Wei, Appl. Surf. Sci., 2015, 346, 335–341.
54. X. Zhang, Q. Huang, M. Liu, J. Tian, G. Zeng, Z. Li, K. Wang, Q. Zhang, Q. Wan and F. Deng, Appl. Surf. Sci., 2015, 343, 19–27.
55. X. Zhang, G. Zeng, J. Tian, Q. Wan, Q. Huang, K. Wang, Q. Zhang, M. Liu, F. Deng and Y. Wei, Appl. Surf. Sci., 2015, 351, 425–432.
56. X. Zhang, S. Wang, L. Xu, L. Feng, Y. Ji, L. Tao, S. Li and Y. Wei, Nanoscale, 2012, 4, 5581–5584.
57. M. Liu, J. Ji, X. Zhang, X. Zhang, B. Yang, F. Deng, Z. Li, K. Wang, Y. Yang and y. wei, J. Mater. Chem. B, 2015, 3, 3476–3482.
58. X. Zhang, M. Liu, Y. Zhang, B. Yang, Y. Ji, L. Feng, L. Tao, S. Li and Y. Wei, RSC Adv., 2012, 2, 12153–12155.
59. X. Zhang, J. Ji, X. Zhang, B. Yang, M. Liu, W. Liu, L. Tao, Y. Chen and Y. Wei, RSC Adv., 2013, 3, 21817–21823.
60. Q. Wan, M. Liu, J. Tian, F. Deng, Y. Dai, K. Wang, Z. Li, Q. Zhang, X. Zhang and Y. Wei, RSC Adv., 2015, 5, 38316–38323.
61. Q. Wan, J. Tian, M. Liu, G. Zeng, Z. Li, K. Wang, Q. Zhang, F. Deng, X. Zhang and Y. Wei, RSC Adv., 2015, 5, 25329–25336.
62. M. Sun, B. Ding and J. Yu, RSC Adv., 2013, 2, 1373–1378.
63. S. Yang, J. Li, D. Shao, J. Hu and X. Wang, J. Hazard. Mater., 2009, 166, 109–116.

---

Footnote

† Electronic supplementary information (ESI) available: TEM images and TGA analysis of CNT samples are provided in the supplementary information. See DOI: 10.1039/c5ra08908e

---