N-doped carbon xerogels as adsorbents for the removal of heavy metal ions from aqueous solution

Abstract

Comparative studies of the textural properties of carbon xerogel (CX) and commercial activated carbon (AC) demonstrated that CX has higher total pore volume and pore diameter, whereas its adsorption capacity for Pb(II) ions in aqueous solution is lower. Herein, the purpose of the present study is to improve the adsorption performance by the introduction of N into the CX matrix on the basis of the extraordinary textural property. Accordingly, the surface chemistry, porous texture, and morphology of the obtained N-doped CX (NCX) were modified. The removal efficiencies of Pb, Zn and Cu by NCX were 1.97, 1.67, and 1.64 times of those by CX, respectively. It was found that the surface density of adsorbed metal ions per unit specific surface area of NCX increased as a linear function with the increase in the N content. The adsorption process of Pb(II) ions followed the pseudo-second-order kinetics and Langmuir model with a maximum adsorption capacity of 83.8 mg g⁻¹, indicating that NCX would be a promising adsorbent for removal of heavy metal ions from water.

---

1. Introduction

Wastewater discharged from the fertilizer and pesticide industry, fuel production industry, mining industry and electroplating process is characterized as having a high content of heavy metals.¹ Since they cannot be degraded biologically like organic pollutants, they are considered as one of the most hazardous substances to human public health owing to the accumulative effect in the human body.² For instance, Pb-poisoning causes severe damage to the neurological system, brain, liver, kidney and reproductive system,^3,4 Cu leads to Wilson's disease, mucosal irritation, hepatic and rental damage, central nervous problem,⁵ and Zn-poisoning leads to vomiting, stomach cramps, diarrhea, and nausea.⁶ Therefore, extensive efforts have been devoted to developing technologies for removing heavy metals from water such as chemical precipitation,⁷ ion exchange,⁸ reverse osmosis,⁹ coagulation,¹⁰ nanofiltration,¹¹ and adsorption.¹² However, some of methods suffer from drawbacks such as incomplete removal,⁷ high capital and operating costs,⁹ high chemical reagent,⁷ generation of toxic sludge,¹⁰ or membrane scaling,¹¹ fouling and blocking.^9,13 Adsorption process is regarded as one of the most effective techniques because of its simplicity, low cost and high efficiency.^14,15

Various adsorbents have been used for the removal of heavy metals from aqueous solution including activated carbon, carbon nanotube sheets, graphene oxide, biosorbents, zeolites, clays, metal oxides, chitosan hydrogel beads and acidic soils.^2,16–23 Among of them, carbon-based adsorbents are especially known because of their inertness to surrounding environment, mechanical stability and high porous structure with specific surface chemical properties.^24–27 In recent years, carbon xerogel (CX) has been attracted a considerable attention due to the three dimensional network in the texture with relatively high porosities.^28,29 The existence of micropores and mesopores is the origin of the utilization in the adsorption of organic contaminants from aqueous solution.^24,30–33 However, few studies on the removal of heavy metals by CX were conducted except for Girgis and coworkers.³⁴ It was demonstrated that the adsorption uptake of Cu(II) ions onto CX using single bottle test was in the range of 32–130 mg g⁻¹, which were controlled by the specific surface area and acidic O-functional groups on surface. They addressed that CX is a new, nanoporous type of porous carbon adsorbents and would be promising for the removal of heavy metals in wastewater. The physical adsorption is dominant since the quantities of acidic O-functional groups, resulting in the chemical adsorption, are low. Incorporating the oxygen or nitrogen functional groups on the material surface in order to enhance the adsorption capacity for heavy metal ions is one of the most simply and efficient method.^35,36 An intensive study of the different N-containing organic compounds on N-doped carbon xerogels (NCX) were conducted by Pérez-Cadenas et al.²⁹ The use of melamine yielded the organic xerogels with a higher N content due to the high hydrophobic character of the amino groups. However, these workers did not concentrate on the adsorption study of heavy metal ions in aqueous solution.

In the present work, the comparative studies on the textural properties of CX and the commercial activated carbon (AC) were firstly conducted and their corresponding adsorption capacities on Pb(II) ions were investigated. The surface chemistry and porous texture of NCX were tuned induced by the introduction of melamine into CX in order to improve higher adsorption capacity. The adsorption performances on heavy metal ions involving Pb(II), Zn(II) and Cu(II) were evaluated and their correlations with the N contents in adsorbents were further determined. Moreover, the adsorption kinetics and isotherms were studied in order to clarify the adsorption mechanism of heavy metals in water onto NCX.

2. Experimental

2.1. Synthesis of N-doped carbon xerogels

The synthesis of CX was followed by Pekala's method,³⁷ and melamine (>98%, TCI) acted as N source for the preparation of NCX. The organic xerogels were obtained from the polycondensation of resorcinol (>99%, TCI) and formaldehyde (37%) using sodium carbonate as the catalyst. Melamine was blended together into the precursor solution if desired. Three molar ratios of resorcinol to catalyst (R/C) were set at 100, 150, and 200, and the corresponding amounts of catalyst were 0.0876 g, 0.0584 g, and 0.0438 g, respectively, where 9.09 g of resorcinol (R) was used. The amount of melamine ranged from 1 g to 2 g. The molar ratio of resorcinol to formaldehyde (R/F) is 0.5 and this value was fixed for all synthesis. The obtained organic gels were carbonized in the furnace at 800 °C with the heating rate of 2 °C min⁻¹ for 12 h where the flow rate of N₂ gas was set at 100 mL min⁻¹. Girgis et al. demonstrated that higher heat treatment temperature could result in higher porosity,³⁴ and the obtained high porosity should be beneficial to the N-doping. The synthesis scheme of NCX is shown in Fig. 1. A simple notation of samples was used in the following context. For example, the sample of NCX-150-2 represents that the molar ratio of R/C is 150 and 2 g melamine was used in the recipe.

Fig. 1 The synthesis scheme of N-doped carbon xerogel.

2.2. Characterization of N-doped carbon xerogels

Surface morphologies of N-doped carbon xerogel were examined by the Field Emission Scanning Electron Microscopy (FE-SEM) (CorlzeisD, ultra 55). The specific surface areas and textural properties were determined by nitrogen isotherms at 77 K on surface area analyzer (Quantachrome, ASIC-2). The Brunauer–Emmett–Teller (BET) method was used for the calculation of the surface areas, while the t-plot method was used to provide the area and volume of micropores (V_micro). The pore diameter (D_pore) was determined using the Barrett–Joyner–Halenda (BJH) pore size distribution method. Elemental analysis was performed with an elemental analyzer (Flash EA 1112).

The point of zero charge (pH_pzc) of CX and NCX was determined by the following steps. Firstly, 20 mL of 0.01 M NaCl solution filled in different flasks, and the solution pH was adjusted by 0.1 M NaOH or HCl in the range from 3 to 11. Secondly, 20 mg adsorbents added into each flask and then were shaken in an electric shaker for 12 h to reach equilibrium. Blank tests were also carried out for comparison. Finally, the pH values of samples and blank were measured. The pH_pzc value is the point where the curve of the pH value of blank crosses with that of sample as reported.

2.3. Batch mode adsorption studies of heavy metal ions

The required amounts of heavy metal salts were dissolved in 1 L distilled water, i.e. Pb(NO₃)₂, Cu(NO₃)₂ and Zn(NO₃)₂ (Primary Reagent, Aladdin). The initial pH of solution was adjusted with 0.1 M HNO₃. In each batch adsorption experiment, 50 mg adsorbents were added into the bottles with 50 mL heavy metal solution, and the initial pH of solution was set at 4.5 if not specified. The test bottles were shaken at 150 rpm in an electric shaker for 12 h at 25 °C to reach equilibrium. Afterwards, the samples were withdrawn and filtered with a 0.22 μm syringe filter. The residual concentration of metal ions in solution was analyzed by Atomic Absorption Spectroscopy (AAS) (Shimadzu, AA-6800). The commercial activated carbon (AC) purchased from Jiangsu Xushui Company, China, was occupied for the comparative study of the adsorption capacity.

The removal efficiency of heavy metals in solution was calculated as follows:

where C₀ and C_t are the concentration of heavy metals in solution at the initial and any time t (mg L⁻¹), respectively. The adsorbed amount of heavy metal by per gram of adsorbent, q_m (mg g⁻¹), was calculated by:where V is the volume of the treated heavy metal solution (L), and m is the mass of the adsorbent added into the solution (g). C_e is the equilibrium concentration of heavy metals in solution (mg L⁻¹) for 12 h in the present work. Before that point of time, the concentration at any time t is represented as C_t.

The surface density of adsorbed amounts of heavy metals at equilibrium to per unit specific surface area, q_s (mg m⁻²), was evaluated according to the method recommended by Girgis and coworkers.³⁴

where S_BET is the specific surface area of adsorbent (m² g⁻¹).

3. Results and discussion

3.1. Carbon xerogel for the removal of Pb(II) ions

The comparative study on adsorption of Pb(II) ions in water by CX and AC was conducted under the initial concentration ranged from 100 mg L⁻¹ to 400 mg L⁻¹, which is representative for industrial wastewater.² The adsorption capacities of q_m were plotted against the initial concentrations, as shown in Fig. 2. It is observed that q_m of both adsorbents increased rapidly with the increasing of the initial concentration and finally obtained the comparatively steady values (or low gradient of increasing), being 31.3 mg g⁻¹ and 38.6 mg g⁻¹ for CX and AC, respectively. Apparently, the q_m of CX is significantly lower than that of AC regardless of initial concentration. The textural properties of the obtained CX and commercial AC are listed in Table 1. The BET surface area of CX is 605.9 m² g⁻¹, which is a little lower than that of AC. However, the surface area of micropore is only 55.5% of AC, which may be the possible reason for the lower adsorption capacity of CX. For the other reason, the amounts of functional groups such as acidic O-functional groups on the CX surface may be low. It is because that the O-functional groups would be decomposed in the high heat treatment temperature used in the synthesis procedure.² Moreover, higher heat treatment temperature can result in higher porosity. Herein, the surface area and total pore volume of CX obtained by the present work are much higher than that by Girgis et al. in the same R/C value of 200.³⁴ The total pore volume and pore diameter of the obtained CX are 1.48 cm³ g⁻¹ and 9.79 nm, respectively, almost 1.7 times and 3.2 times higher than that of AC. It is the extraordinary textural property in the CX that would be beneficial to the element-doping for the modification of the surface chemistry, which is expected to improve the adsorption performance.

Fig. 2 Adsorption performance of CX and AC for Pb(II) ion in water.

Table 1 The textural properties of CX and AC

Samples	SBET (m^2 g^−1)	Vtotal (cm^3 g^−1)	Smicro (m^2 g^−1)	Vmicro (cm^3 g^−1)	Dpore (nm)
CX	605.9	1.48	319.2	0.24	9.79
AC	643.0	0.55	575.0	0.27	2.33

---

3.2. Characterization of N-doped carbon xerogels

3.2.1 Physical properties of N-doped carbon xerogels. The textural properties of the synthesized N-doped CX with different amounts of catalyst and melamine are shown in Table 2. It is observed that the BET surface areas of NCX ranged from 530.0 m² g⁻¹ to 608.8 m² g⁻¹, which was slightly reduced in comparison with that of CX. The total pore volume and pore diameter were significantly reduced when more or less addition of catalyst were occupied. The worse properties were obtained by NCX-100-2 and NCX-200-2. The total pore volumes of NCX-100-2 and NCX-200-2 are 0.45 cm³ g⁻¹ and 0.50 cm³ g⁻¹, being only 30.4% and 33.8% of CX, respectively. NCX-150-2 achieved the highest total pore volume and pore diameter, being 1.15 cm³ g⁻¹ and 8.02 nm, respectively. Therefore, the R/C value of 150 in this recipe was optimal based on the combined consideration of surface area and total pore volume. It is noticed that the total pore volume and pore diameter significantly increased with the increasing of the addition of melamine. It may be because that the use of melamine could produce a smaller particle diameter, and herein the texture with higher porosity was formed.²⁹

Table 2 Textural properties of the synthesized N-doped CX

Samples	SBET (m^2 g^−1)	Vtotal (cm^3 g^−1)	Smicro (m^2 g^−1)	Vmicro (cm^3 g^−1)	Dpore (nm)
NCX-100-2	531.0	0.45	196.2	0.16	3.41
NCX-150-2	573.3	1.15	255.1	0.20	8.02
NCX-200-2	530.0	0.50	273.4	0.21	3.75
NCX-150-1	560.4	0.57	197.7	0.16	4.21
NCX-150-1.5	608.8	0.93	246.5	0.20	6.38

---

The NCX surface morphologies with both R/C ratios are shown in Fig. 3, where CX was listed in comparison. It is observed that NCX-150-2 exhibits a rough and highly porous structure with the spherical-like interconnected particles. A little bit aggregation of particles were formed due to the doping of melamine. However, NCX-200-2 suffered from the severe cluster of nanoparticles due to low addition of catalyst, which would result in low specific surface area and pore volume, corroborating the earlier results. It reveals that the addition of melamine did not significantly change the surface morphologies of CX when the proper amounts of added catalyst were occupied in the synthesis.

Fig. 3 The surface morphologies of (a) CX, (b) NCX-150-2 and (c) NCX-200-2.

3.2.2 Surface chemical properties of N-doped carbon xerogels. The data obtained from elemental analysis of the global composition together with the raw addition of nitrogen (N_raw) are summarized in Table 3. The overall obtained N content in NCX is in the range from 2.19% to 3.29%. It is expected that high quantity of melamine addition would result in high N content. The highest overall N content of 3.29% was achieved by NCX-100-2, although its N yield is lower than those of NCX-150-1 and NCX-150-1.5. It indicated that the N yields would be low when the amounts of added melamine were high. The N losses are in the range from 20.3% to 45.8% during the carbonization owing to the heat treatment. The obtained N content of NCX-200-2 is lower than that of either NCX-150-2 or NCX-100-2, indicating that the losses increased with the increase of R/C value due to the shortage of catalyst.

Table 3 Elemental composition of N-doped carbon xerogels

Samples	Nraw (%)	C (%)	H (%)	N (%)	N yield (%)
NCX-100-2	5.26	85.51	1.28	3.29	62.5
NCX-150-2	5.27	83.51	1.29	2.97	56.4
NCX-200-2	5.27	81.95	1.41	2.86	54.2
NCX-150-1	2.74	85.69	1.02	2.19	79.7
NCX-150-1.5	4.03	83.66	1.04	2.80	69.5

---

XPS analysis was further performed for better understanding of the N doping chemistry of CX. The N_1s core-level spectra of NCX-150-2 are shown in Fig. 4. The banding energies of the peaks were assigned to different N components in accordance with data in the literature, and the percentage of each N component is in accordance with the corresponding peak areas.^29,38–41 The distinct fitting peaks at around 398.0 eV, 399.5 eV, and 401.0 eV can be devolved to pyridinic-N (N-6), pyrrolic-N (N-5), and graphitic N (N-Q), respectively, which means that the different forms of N atoms substituted for C atoms in flake like layer. It determined that the percentage of N-Q is around 49.2%. The dominated graphitic N in NXC may contribute two lone-pair electrons to bind heavy metal ions. Moreover, heavy metals acting as soft Lewis acids have a high affinity to the amino groups induced by ion-exchange process.^12,42,43

Fig. 4 N_1s core-level spectra of N-doped carbon xerogel.

Some researchers have investigated the reaction formation mechanism of NCX and the corresponding N functionalities.^44–46 They addressed that the reaction of resorcinol and melamine with formaldehyde leaded to hydroxymethylation at first, whereas hydrogen atoms in the benzene ring and NH₂ groups were substituted by methylol groups (–CH₂OH), followed by polycondensation of the resulting hydroxymethyl resorcinol and methylolmelamine. Depending on the degree of crosslinking, a couple of different bonds can appear, such as methylene bridges (=N–CH₂–N=), amino groups (–NH₂), –NR₃, ether linkages (–CH₂–O–CH₂–), –CH–CH₂–, and hydroxyl groups (CH₂–OH). The complex surface N functionalities can be obtained after pyrolysis.

The pH_pzc determination of CX and NCX was described in Fig. 5. The obtained pH_pzc of CX and NCX-15-2 are 6.41 and 7.61, respectively. The acidity of CX may be derived from the acid O-containing functional groups on the surface, and the surface chemical properties of NCX were tuned to the alkalinity due of amino functional groups.

Fig. 5 The pH_pzc determination of CX and NCX.

3.3. N-doped CX for removal of heavy metal ions

3.3.1 Adsorption performance of N-doped CX. The removal efficiencies of heavy metals including Pb, Zn and Cu for the initial concentration of 50 mg L⁻¹ on NCX are shown in Fig. 6, where CX was listed for comparison. The significant increases were observed by NCX for each heavy metal. NCX-150-2 obtained the highest efficiency, and the corresponding removal efficiencies of Pb, Zn and Cu are 74.4%, 80.3%, and 83.1%, respectively, which are 1.97, 1.67, and 1.64 times of those by CX. NCX-150-1 was expected to obtain the lowest removal efficiency among of NCX due to the lowest N content. For 2 g addition of melamine, NCX-200-2 obtained lower efficiency not only for worse textural property but also for lower N content. The efficiencies by NCX-200-2 are a bit higher than those by NCX-150-1 as well, although its total specific surface area and pore diameter are lower. It may be because of the relatively higher N content in NCX-200-2. Similarly, the efficiency by NCX-150-1.5 is lower than that by NCX-100-2, although it obtained the highest BET surface area and relatively high pore diameter. It is demonstrated that the adsorption performance of N-doped CX is mainly dependent on the surface N functional groups rather than the textural properties. Therefore, it indicates that N-doped CX are simple in synthesis and high in efficiency for the removal of heavy metals. Additionally, the removal efficiency of Cu is slightly higher than those of Pb and Zn. The differences on the adsorption uptake of heavy metals may be caused by differences in hydration energies relative with the diameter of the hydrated cation and its water molecules in the hydration shell.⁴²

Fig. 6 Removal efficiencies of heavy metals by NCX.

3.3.2 Correlation between the N content and surface density. The surface densities of adsorbed heavy metals at equilibrium to per unit specific surface area of NCX were plotted against the N content, as shown in Fig. 7. It is observed that the surface densities increased as a linear function with the increase in the N content. It reveals that the surface chemistry property induced by the N doping may be the predominant contributor to the adsorption process rather than the total specific surface area. The slopes of fitting line for Pb, Zn and Cu are 0.0114, 0.0102, and 0.0107, respectively, while the correlation coefficients of R² are 0.992, 0.996, and 0.998. Higher slope value was achieved by Pb, indicating that its adsorption uptake can be more easily improved induced by the N-doping in comparison with those of Zn and Cu.

Fig. 7 Correlation between the N content and surface density.

3.3.3 Adsorption isotherms. The adsorption equilibrium isotherms of Pb with the initial concentration from 30 mg L⁻¹ to 400 mg L⁻¹ on NCX were further investigated. The Henry, Langmuir, Freundlich models were used for fitting the adsorption data, whose corresponding equations are expressed as follows, respectively.

q_e = kC_e (4)

The parameter and correlation coefficients by various carbon xerogels for the Langmuir model are summarized in Table 4 and those for Henry and Freundlich models are in Table S1,† where the results by CX were listed for comparison. The obtained correlation coefficients of R² only for the Langmuir model are higher than 0.9, indicating that the experimental data fitted well into the Langmuir isotherm. The calculated maximum adsorption capacity, q_max, followed the order NCX-150-2 > NCX-100-2 > NCX-150-1.5 > NCX-200-2 > NCX-150-1 > CX, being 83.8 mg g⁻¹, 74.5 mg g⁻¹, 73.2 mg g⁻¹, 71.7 mg g⁻¹, 69.9 mg g⁻¹, and 64.2 mg g⁻¹, respectively.

Table 4 Parameters and correlation coefficients for the isotherm models

Models	Parameter	CX	NCX-100-2	NCX-150-2	NCX-200-2	NCX-150-1	NCX-150-1.5
Langmuir	b	0.015	0.0363	0.0278	0.0247	0.0479	0.0650
	qm	64.2	74.5	83.8	71.7	69.9	73.2
	R^2	0.999	0.996	0.987	0.994	0.962	0.979
D-R	E	8.84	15.65	14.07	12.14	11.64	11.89
	R^2	0.887	0.946	0.956	0.957	0.849	0.946

---

Furthermore, Dubinin-Radushkevich model was determined in order to study the chemical and physical behavior in the surface adsorption process.⁴⁷ The linear form of D-R equation is expressed as follow.

ln q_e = ln q^D-R_m − βε² (7)

E = (−2β)^−1/2 (9)

where q^D-R_m represents the D-R adsorption capacity (mmol g⁻¹), β is the constant related to the adsorption energy (mol² kJ⁻²) and ε is the Polanyi potential, R is the gas constant (8.314 J mol⁻¹ K⁻¹) and T is the experimental temperature (K). The determination of mean adsorption energy E (kJ mol⁻¹) can be calculated from β. If E < 8 kJ mol⁻¹, it represents that the adsorption process is governed by physisorption. If 8 kJ mol⁻¹ < E < 16 kJ mol⁻¹, the chemisorption dominates the adsorption process.⁹

The calculated values of E from D-R model are as well summarized in Table 4, and the obtained correlation coefficients of R² are close to 0.9, indicating that the experimental data fitted well into the D-R isotherm as well. It is found that the E values for NCX are much higher than 8 kJ mol⁻¹ and follow the orders: NCX-100-2 > NCX-150-2 > NCX-200-2 > NCX-150-1.5 > NCX-150-1. It should be emphasized that these orders are consistent with the obtained N contents in NCX, and it reveals that the chemisorption reactions induced by the N functional groups may be predominant in the adsorption process by NCX. The E value obtained by NCX-100-2 is 15.65 kJ mol⁻¹, which is 1.77 times than that by CX. The E value for CX is 8.84 kJ mol⁻¹, which is slightly higher than 8 kJ mol⁻¹, indicating that chemisorption with the contributions of some acidic O-containing groups occurred in the adsorption process.

3.4. Effect of initial pH on removal efficiencies

Fig. 8 shows the effect of initial pH on the removal efficiencies of Pb for CX and NCX-150-2. The efficiencies increased significantly with the increase in the initial pH from 2 to 7.5, either for CX or NCX-150-2. The removal efficiency by NCX-15-2 is only 9.4% at pH 2, and while it reached 97.7% at pH 7.5. At low pH, the relatively high concentration of H⁺ would induce highly positive surface charge and compete the binding sites of adsorbent with Pb ions. Hence, the increase of the repulsive forces between the adsorbent and positive Pb ions may result in low removal efficiency. However, at high pH, especially at pH > pH_pzc, the surface charge of adsorbent would be negative and the electrostatic interactions between the metal ions and adsorbent became stronger, which is contributed to increasing the removal efficiency. Besides, some precipitation of Pb will be formed when the solution pH reached 7.5.

Fig. 8 Effect of initial pH on the removal efficiencies of Pb.

The ion-exchange process of metal ions not only occurred on the CX surface with the acidic O-containing groups but also on NCX with the basic N-containing groups. Moreover, the surface complexation on NCX was also the contributor for the adsorption of heavy metals. The surface adsorption reactions can be described as follows, where M is the metal ions.

2(CX-OH) + M²⁺ → (CX-O)₂M + 2H⁺ (10)

2(CX-NH)⁺ + M²⁺ → (CX-N)₂M + 2H⁺ (11)

CX-N + M²⁺ → [M → N-CX]²⁺ (12)

It infers that the ion-exchange process would induce the decreasing of the solution pH due to the release of H⁺ ions other than the surface complexation. The measured final pH values were plotted against the initial pH, as shown in Fig. 9. It is observed that the solution pH was stable at low initial pH because the high concentration of H⁺ may inhibit the ion-exchange process. The final pH decreased with the increase in the initial pH where the contribution of ion-exchange process improved. Lower final pH was obtained by CX, indicating that higher amounts of H⁺ ions were released. However, it should be noticed that the adsorbed uptakes of Pb on NCX are much higher than those on CX. It further demonstrated that the chemisorption process of metals ions by NCX proceeded in the surface complexation other than the ion-exchange process.

Fig. 9 Relationship of the measured final pH with initial pH of treated solution.

3.5. Adsorption kinetics

The adsorption kinetic studies were conducted to determine the equilibrium time for the adsorption of Pb with various carbon xerogels. The results are presented in Fig. 10(a). The removal efficiency by NCX-150-2 was 63.5% after 1 h, in contrast, those of NCX-150-1 and CX were 53.2% and 32.3%, respectively. The adsorption by NCX-150-2 achieved equilibrium gradually at about 3 h and the removal efficiency reached 74.4%. It is as well observed that the adsorption rate by CX seems faster than that of NCX.

Fig. 10 The removal efficiencies against the contact time (a) and adsorption pseudo-second-order kinetic model (b).

In order to give an in-depth insight into the adsorption mechanism, three kinetic models including the intraparticle diffusion, the pseudo-first-order, and the pseudo-second-order models were applied to fit the experimental data obtained from the adsorption kinetic experiments of Pb ions onto the typical samples. The descriptions of intraparticle diffusion and the pseudo-first-order modes are presented in the ESI† and the fitting curves for are shown in Fig. S1,† and the corresponding parameters and correlation coefficients are listed in Table S2.† The pseudo-second-order kinetic model assumes that the maximum adsorption corresponds to a saturated monolayer of metal ions onto the adsorbent surface.³⁵

3.5.1 Pseudo-second-order model. where q_t is the adsorbed uptake of Pb at contact time t. k₂ is the rate constant. The linear fit of t/q_t against t can be described in Fig. 10(b).

The parameters and correlation coefficients of the fitting curves for pseudo-second-order modes are listed in Table 5. It is observed that the fitting curves of pseudo-second-order model seemed quite well with the experimental data in comparison with those of intraparticle diffusion and the pseudo-first-order modes, and all obtained correlation coefficients R² are close to 0.99. The calculated values of q_e for CX, NCX-150-2, and NCX-150-1 are 18.9 mg g⁻¹, 37.6 mg g⁻¹, and 32.4 mg g⁻¹, respectively. The adsorption capacities of NCX were significantly increased in comparison with CX. However, its rate constant is much lower than that of CX. It is because that the total pore volume of NCX is lower and the diffusion process would be restricted which decreased the adsorption rate.

Table 5 Parameters and correlation coefficients for the kinetic models

Models	Parameter	CX	NCX-150-2	NCX-150-1
Pseudo-second-order	k2 (g mg^−1 h^−1)	0.398	0.213	0.153
	qe (mg g^−1)	18.9	37.6	32.4
	R^2	0.999	0.999	0.998

---

3.6. Adsorption thermodynamics

The influences of operating temperature on the adsorption thermodynamic were carried out at the temperatures from 288 K to 313 K. The thermodynamic parameters including the Gibbs free energy change (ΔG), enthalpy change (ΔH), and entropy change (ΔS) were calculated using following equations:

ΔG = −RT ln K_c (14)

ΔG = ΔH − TΔS (15)

where K_c is the equilibrium constant. C_a and C_e are the Pb ion concentrations (mg L⁻¹) on the adsorbent and in the solution at equilibrium, respectively. The plots of log K_c versus 1/T are presented in Fig. 11. ΔH and ΔS were calculated from the slope and the intercept of the plots of log K_c. All the values and coefficients are listed in Table 6. It shows that the value of ΔH is positive, indicating that an endothermic process occurs in the adsorption of heavy metals onto NCX. Herein, the removal efficiencies increased with the increase in the operating temperature. It can be explained by either the increase of active surface sites available for adsorption or the decreasing of the boundary layer thickness, namely the decreasing of the mass transfer resistance of metal ions.¹ The high values of ΔH in the range from 46.4 kJ mol⁻¹ to 65.9 kJ mol⁻¹ indicated the strong interaction between the metal ions and adsorbents. The adsorption process was of the spontaneous nature since the value of ΔG is negative.

Fig. 11 Plots of the equilibrium constant at different temperature for the adsorption process.

Table 6 Thermodynamic parameters for the adsorption process

	T (K)	Kc	ΔG (kJ mol^−1)	ΔS (J mol^−1 K^−1)	ΔH (kJ mol^−1)	R^2
NCX-100-2	298	2.86	−2.64	0.165	46.4	0.996
	308	5.56	−4.29
	313	7.03	−5.11
	318	9.35	−5.94
NCX-150-2	298	2.91	−2.61	0.193	54.8	0.998
	308	5.79	−4.54
	313	8.06	−5.50
	318	11.83	−6.46
NCX-200-2	298	1.88	−1.35	0.182	53.0	0.852
	308	3.23	−3.17
	313	3.81	−4.08
	318	8.14	−4.99
NCX-150-1	298	1.80	−1.60	0.227	65.9	0.974
	308	5.23	−3.87
	313	6.56	−5.00
	318	9.75	−6.14

---

4. Conclusions

N-doped carbon xerogels were successfully synthesized by using melamine as the N source for improving higher adsorption capacity of NCX. The removal efficiencies of Pb, Zn and Cu by NCX-150-2 were 74.4%, 80.3%, and 83.1%, respectively, which were 1.97, 1.67, and 1.64 times of those by CX. The excellent adsorption performance of N-doped CX is mainly dependent on the surface N functional groups rather than the textural properties. The obtained percentage of N content in NCX was in the range from 2.19% to 3.29%. The surface density of adsorbed metal ions to per unit specific surface area of NCX increased as a linear function with the increase in the N content. The obtained pH_pzc of NCX-150-2 is 7.61, and the alkalinity of NCX induced by amino functional groups would be beneficial to the adsorption process of metal ions. The surface complexation with the lone pair electrons offered by the N atom may be predominant rather than ion-exchange process. The adsorption process of Pb(II) ions followed the pseudo-second-order kinetics and Langmuir model with the maximum adsorption capacity of 83.8 mg g⁻¹, indicating that NCX would be a promising adsorbent for the removal of heavy metal ions from water.

Acknowledgements

This work was partially supported by the National Natural Science Foundation of China (NSFC) (no. 2140197, U1462201 and 21406044), and the Zhejiang Postdoctoral Sustentation Fund (no. BSH1302042).

References

1. A. K. Meena, G. K. Mishra, P. K. Rai, C. Rajagopal and P. N. Nagar, Removal of heavy metal ions from aqueous solutions using carbon aerogel as an adsorbent, J. Hazard. Mater., 2005, 122(1–2), 161–170.
2. M. A. Tofighy and T. Mohammadi, Adsorption of divalent heavy metal ions from water using carbon nanotube sheets, J. Hazard. Mater., 2011, 185(1), 140–147.
3. H. Gurer-Orhan, H. U. Sabır and H. Özgüneş, Correlation between clinical indicators of lead poisoning and oxidative stress parameters in controls and lead-exposed workers, Toxicology, 2004, 195(2–3), 147–154.
4. B. C. Campbell, H. L. Elliott and P. A. Meredith, Lead exposure and renal failure: Does renal insufficiency influence lead kinetics?, Toxicol. Lett., 1981, 9(2), 121–124.
5. E. I. El-Shafey, M. Cox, A. A. Pichugin and Q. Appleton, Application of a carbon sorbent for the removal of cadmium and other heavy metal ions from aqueous solution, J. Chem. Technol. Biotechnol., 2002, 77(4), 429–436.
6. U. Ipek, Removal of Ni(II) and Zn(II) from an aqueous solution by reverse osmosis, Desalination, 2005, 174(2), 161–169.
7. F. Fu, L. Xie, B. Tang, Q. Wang and S. Jiang, Application of a novel strategy-Advanced Fenton-chemical precipitation to the treatment of strong stability chelated heavy metal containing wastewater, Chem. Eng. J., 2012, 189–190, 283–287.
8. W. Yang, D. Wu and R. Fu, Adsorption of theophylline from aqueous solution on organic aerogels and carbon aerogels, J. Porous Mater., 2008, 16(5), 507–512.
9. H. Ozaki, K. Sharma and W. Saktaywin, Performance of an ultra-low-pressure reverse osmosis membrane (ULPROM) for separating heavy metal: effects of interference parameters, Desalination, 2002, 144(1–3), 287–294.
10. A. Örnek, M. Özacar and İ. A. Şengil, Adsorption of lead onto formaldehyde or sulphuric acid treated acorn waste: Equilibrium and kinetic studies, Biochem. Eng. J., 2007, 37(2), 192–200.
11. A. G. El Samrani, B. S. Lartiges and F. Villiéras, Chemical coagulation of combined sewer overflow: heavy metal removal and treatment optimization, Water Res., 2008, 42(4–5), 951–960.
12. Y. F. Jia, B. Xiao and K. M. Thomas, Adsorption of Metal Ions on Nitrogen Surface Functional Groups in Activated Carbons, Langmuir, 2001, 18(2), 470–478.
13. K. Y. Kang, B. I. Lee and J. S. Lee, Hydrogen adsorption on nitrogen-doped carbon xerogels, Carbon, 2009, 47(4), 1171–1180.
14. A. Agrawal, K. K. Sahu and B. D. Pandey, Systematic studies on adsorption of lead on sea nodule residues, J. Colloid Interface Sci., 2005, 281(2), 291–298.
15. A. Demirbas, Heavy metal adsorption onto agro-based waste materials: A review, J. Hazard. Mater., 2008, 157(2–3), 220–229.
16. O. S. Pokrovsky, A. Probst, E. Leviel and B. Liao, Interactions between cadmium and lead with acidic soils: experimental evidence of similar adsorption patterns for a wide range of metal concentrations and the implications of metal migration, J. Hazard. Mater., 2012, 199–200(0), 358–366.
17. W. L. Yan and R. Bai, Adsorption of lead and humic acid on chitosan hydrogel beads, Water Res., 2005, 39(4), 688–698.
18. K. Zhang, W. H. Cheung and M. Valix, Roles of physical and chemical properties of activated carbon in the adsorption of lead ions, Chemosphere, 2005, 60(8), 1129–1140.
19. V. Singh, A. K. Sharma, P. Kumari and S. Tiwari, Efficient chromium(VI) adsorption by Cassia marginata seed gum functionalized with poly(methylmethacrylate) using microwave irradiation, Ind. Eng. Chem. Res., 2008, 47(15), 5267–5276.
20. Y. S. Ok, J. E. Yang, Y.-S. Zhang, S.-J. Kim and D.-Y. Chung, Heavy metal adsorption by a formulated zeolite-Portland cement mixture, J. Hazard. Mater., 2007, 147(1–2), 91–96.
21. R. Celis, M. C. HermosÍn and J. Cornejo, Heavy Metal Adsorption by Functionalized Clays, Environ. Sci. Technol., 2000, 34(21), 4593–4599.
22. H. Tamura, N. Katayama and R. Furuichi, Modeling of Ion-Exchange Reactions on Metal Oxides with the Frumkin Isotherm. 1. Acid–Base and Charge Characteristics of MnO₂, TiO₂, Fe₃O₄, and Al₂O₃ Surfaces and Adsorption Affinity of Alkali Metal Ions, Environ. Sci. Technol., 1996, 30(4), 1198–1204.
23. G. Zhao, X. Ren, X. Gao, X. Tan, J. Li, C. Chen, Y. Huang and X. Wang, Removal of Pb(II) ions from aqueous solutions on few-layered graphene oxide nanosheets, Dalton Trans., 2011, 40(41), 10945–10952.
24. S. A. C. Carabineiro, T. Thavorn-Amornsri, M. F. R. Pereira, P. Serp and J. L. Figueiredo, Comparison between activated carbon, carbon xerogel and carbon nanotubes for the adsorption of the antibiotic ciprofloxacin, Catal. Today, 2012, 186(1), 29–34.
25. C. He and X. Hu, Anionic dye adsorption on chemically modified ordered mesoporous carbons, Ind. Eng. Chem. Res., 2011, 50(24), 14070–14083.
26. K. Yang and B. S. Xing, Adsorption of Organic Compounds by Carbon Nanomaterials in Aqueous Phase: Polanyi Theory and Its Application, Chem. Rev., 2010, 110(10), 5989–6008.
27. R. Sitko, E. Turek, B. Zawisza, E. Malicka, E. Talik, J. Heimann, A. Gagor, B. Feist and R. Wrzalik, Adsorption of divalent metal ions from aqueous solutions using graphene oxide, Dalton Trans., 2013, 42(16), 5682–5689.
28. C. Lin and J. A. Ritter, Effect of synthesis pH on the structure of carbon xerogels, Carbon, 1997, 35(9), 1271–1278.
29. M. Pérez-Cadenas, C. Moreno-Castilla, F. Carrasco-Marín and A. F. Pérez-Cadenas, Surface chemistry, porous texture, and morphology of N-doped carbon xerogels, Langmuir, 2009, 25(1), 466–470.
30. L. A. Rodrigues, T. M. B. Campos, M. O. Alvarez-Mendes, A. D. R. Coutinho, K. K. Sakane and G. P. Thim, Phenol removal from aqueous solution by carbon xerogel, J. Sol–Gel Sci. Technol., 2012, 63(2), 202–210.
31. M. C. Almazán-Almazán, F. J. López-Domingo, M. Domingo-García, A. Léonard, M. Pérez-Mendoza, J. P. Pirard, F. J. López-Garzón and S. Blacher, Influence of carbon xerogel textural properties on the dynamic adsorption of methyl iodide, Chem. Eng. J., 2011, 173(1), 19–28.
32. B. S. Girgis, A. A. Attia and N. A. Fathy, Potential of nano-carbon xerogels in the remediation of dye-contaminated water discharges, Desalination, 2011, 265(1–3), 169–176.
33. J. Figueiredo, J. Sousa, C. Orge, M. Pereira and J. Órfão, Adsorption of dyes on carbon xerogels and templated carbons: influence of surface chemistry, Adsorption, 2011, 17(3), 431–441.
34. B. S. Girgis, I. Y. El-Sherif, A. A. Attia and N. A. Fathy, Textural and adsorption characteristics of carbon xerogel adsorbents for removal of Cu(II) ions from aqueous solution, J. Non-Cryst. Solids, 2012, 358(4), 741–747.
35. K.-Y. Shin, J.-Y. Hong and J. Jang, Heavy metal ion adsorption behavior in nitrogen-doped magnetic carbon nanoparticles: isotherms and kinetic study, J. Hazard. Mater., 2011, 190(1–3), 36–44.
36. C. A. Bell, S. V. Smith, M. R. Whittaker, A. K. Whittaker, L. R. Gahan and M. J. Monteiro, Surface-Functionalized Polymer Nanoparticles for Selective Sequestering of Heavy Metals, Adv. Mater., 2006, 18(5), 582–586.
37. R. W. Pekala, Organic aerogels from the polycondensation of resorcinol with formaldehyde, J. Mater. Sci., 1989, 24(9), 3221–3227.
38. J. Lahaye, G. Nanse, P. Fioux, A. Bagreev, A. Broshnik and V. Strelko, Chemical transformation during the carbonisation in air and the pyrolysis under argon of a vinylpyridine–divinylbenzene copolymer by X-ray photoelectron spectroscopy, Appl. Surf. Sci., 1999, 147(1–4), 153–174.
39. J. Lu, X. Bo, H. Wang and L. Guo, Nitrogen-doped ordered mesoporous carbons synthesized from honey as metal-free catalyst for oxygen reduction reaction, Electrochim. Acta, 2013, 108(0), 10–16.
40. H. Wang, X. Bo, C. Luhana and L. Guo, Nitrogen doped large mesoporous carbon for oxygen reduction electrocatalyst using DNA as carbon and nitrogen precursor, Electrochem. Commun., 2012, 21(0), 5–8.
41. J. Casanovas, J. M. Ricart, J. Rubio, F. Illas and J. M. Jiménez-Mateos, Origin of the Large N_1s Binding Energy in X-ray Photoelectron Spectra of Calcined Carbonaceous Materials, J. Am. Chem. Soc., 1996, 118(34), 8071–8076.
42. U. Wingenfelder, B. Nowack, G. Furrer and R. Schulin, Adsorption of Pb and Cd by amine-modified zeolite, Water Res., 2005, 39(14), 3287–3297.
43. J. Aguado, J. M. Arsuaga, A. Arencibia, M. Lindo and V. Gascón, Aqueous heavy metals removal by adsorption on amine-functionalized mesoporous silica, J. Hazard. Mater., 2009, 163(1), 213–221.
44. H. F. Gorgulho, F. Gonçalves, M. F. R. Pereira and J. L. Figueiredo, Synthesis and characterization of nitrogen-doped carbon xerogels, Carbon, 2009, 47(8), 2032–2039.
45. J. R. Pels, F. Kapteijn, J. A. Moulijn, Q. Zhu and K. M. Thomas, Evolution of nitrogen functionalities in carbonaceous materials during pyrolysis, Carbon, 1995, 33(11), 1641–1653.
46. D. Long, J. Zhang, J. Yang, Z. Hu, G. Cheng, X. Liu, R. Zhang, L. Zhan, W. Qiao and L. Ling, Chemical state of nitrogen in carbon aerogels issued from phenol–melamine–formaldehyde gels, Carbon, 2008, 46(9), 1259–1262.
47. M. M. Dubinin, The Potential Theory of Adsorption of Gases and Vapors for Adsorbents with Energetically Nonuniform Surfaces, Chem. Rev., 1960, 60(2), 235–241.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c4ra12441c

---