Tuning the Chemical Hardness of Boron Nitride Nanosheets by

Sep 5, 2017 - These works also showed that the hard and soft acids and bases (HSAB) theory ..... The data of adsorption experiment are shown in Sectio...
0 downloads 0 Views 9MB Size
This is an open access article published under an ACS AuthorChoice License, which permits copying and redistribution of the article or any adaptations for non-commercial purposes.

Article http://pubs.acs.org/journal/acsodf

Tuning the Chemical Hardness of Boron Nitride Nanosheets by Doping Carbon for Enhanced Adsorption Capacity Hongping Li,† Siwen Zhu,‡ Ming Zhang,† Peiwen Wu,‡ Jingyu Pang,‡ Wenshuai Zhu,*,‡ Wei Jiang,† and Huaming Li*,† †

Institute for Energy Research and ‡School of Chemistry and Chemical Engineering, Jiangsu University, Zhenjiang 212013, P. R. China S Supporting Information *

ABSTRACT: The chemical hardness of adsorbents is an important physicochemical property in the process of adsorption based on the hard and soft acids and bases (HSAB) theory. Tuning chemical hardness of adsorbents modulated by their concomitants is a promising approach to enhance the adsorptive capacity in principle. In the present work, we report an efficient strategy that the adsorption capacity for aromatic sulfocompounds can be enhanced by tuning the chemical hardness. This strategy is first theoretically explored by introducing C element into the network of hexagonal boron nitride (h-BN) based on a series of model materials (model_xC, x = 1−5). Computational results show that the chemical hardness is reduced after gradually C-doping, which may lead to an enhancement of adsorption capacity according to the HSAB theory. Then, a series of C-doped h-BN materials (BCNx, x = 10−50) were controlled synthesized. All of the as-prepared materials show better adsorption capacities (e.g., 27.43 mg g−1 for BCN-50) than pure h-BN. Experiment results show that the adsorption capacity correlates well with the C content in the BCN-x, which is consistent with the results predicted by theoretical calculation. This strategy may be helpful to rationally design highly efficient adsorbents in separation engineering and may be expanded to similar two-dimensional materials, where the π−π interaction is the dominant driven force.

1. INTRODUCTION Two-dimensional (2D) layered materials are of particular interest from both scientific and application viewpoints due to their unique planar structures and properties.1−4 Recently, because of their possible applications in separation science, some 2D layered materials (e.g., graphene, hexagonal boron nitride (hBN), transition-metal dichalcogenides, graphitic carbon nitride, and layered metal oxides) have attracted considerable interests.5 Especially, hexagonal boron nitride (h-BN), which owns unique properties compared with other 2D layered materials, such as wide-energy band gap,6 electrical insulation,7 and chemical inertness,8 has been used as adsorbent, catalyst, or other functional materials. As a graphene-like material, h-BN shows good adsorption capacity for many adsorbates, including CO2, organic dyes, 9−11 antibiotics, 12 oil, 13,14 and sulfocompounds.8,15,16 Two possible physicochemical properties are employed to interpret the adsorption capacity of h-BN. One is that the unique polarity of BN bonds and the high surface area of h-BN-related nanostructures provide good adsorption properties.3 The other is that the coexisted Lewis acidic sites (B atom) and basic sites (N atom) can afford various types of chemical interaction between h-BN and adsorbates. In recent years, researchers have summarized that various 2D layered materials can be built from B, C, and N elements in a specific stoichiometric ratio.17 It affords a wide space to modify these materials for different applications. To date, most efforts are © 2017 American Chemical Society

devoted to the facile synthesis, energy storage, and photoredox catalysis.6,17,18 In this study, we report a rational design of ternary C-doped h-BN materials (BCN) materials to enhance the adsorption capacity for separation engineering. It is well known that sulfur removal from fuels is an important topic for both chemical industry and environments.19,20 Sulfocompounds (especially aromatic sulfocompounds) in fuels are difficult to be deeply removed by adsorption. To date, deep desulfurization from fuels by h-BN has still achieved limited success despite its feasible modification in both morphology and atomic structures, which promises a potential space to reach the goal. Hence, how to rationally design highly efficient adsorbents based on h-BN is a significant and attractive subject in petroleum chemical industry. In principle, rational design of highly efficient adsorbents should be based on a deep understanding of the adsorption mechanism for BN-based materials. So far, π−π interaction as a dominant driving force has been used to explain the mechanism of aromatic adsorbate to layered graphene surface.10 As a graphene-like material, similar π−π interaction may exist between h-BN and aromatic sulfocompounds. Chao et al. suggested that virtual orbitals of B existed in h-BN, presenting some Lewis acidity, which can serve as an electron acceptor. And Received: June 15, 2017 Accepted: August 11, 2017 Published: September 5, 2017 5385

DOI: 10.1021/acsomega.7b00795 ACS Omega 2017, 2, 5385−5394

ACS Omega

Article

Figure 1. Frontier orbitals of BN, model_xC (x = 1−5). It should be noted that the LUMO orbitals were selected from LUMO + 3 (BN) and LUMO + 6 (model_1C) due to the turbulence of molecular energy levels.

bond lengths of C−B in model_1C are generally longer than those of B−N (e.g., 1.529 Å for C−B and 1.455 Å for N−B, Figure S1). Besides, after gradually C-doping, the bond lengths of C−B become slightly deformed, which ranges from 1.484 to 1.567 Å. The N−B bonds remain almost unchanged after Cdoping as compared to the computational results. All of the values of frontier orbitals, electronegativity, hardness, and energy gap (for concepts definition, see Section 4.1) of model_xC are listed in Table 1. According to frontier orbital theory, HOMO is

the adsorbate (gatifloxacin) has two aromatic rings, which can serve as a π-electron donor.12 Besides, Xiong et al. suggested that Lewis acid−base interaction plays an important role in removing aromatic sulfocompounds from fuels.8 In addition, the mechanism of adsorption of aromatic sulfocompounds is suggested to be π-complexation on other adsorbents, such as Ag-Y and Cu-Y zeolites.21−24 Gao et al. considered that the adsorptive activity of some metal compounds (such as AlCl3 and FeCl3) originates from their strong Lewis acidity and the sulfur basicity of the sulfocompound, which leads to the acid−base complexation.25 These works also showed that the hard and soft acids and bases (HSAB) theory can serve as a fundamental tool to understand the adsorptive strength of aromatic sulfocompounds. Moreover, our previous work also suggests that the adsorptive selectivity correlates well with the hardness of the sulfocompounds.26 Previous works have reported that aromatic sulfocompounds (e.g., dibenzothiophene, DBT) are soft Lewis bases.25,27−31 On the basis of the HSAB principle, it is known that hard acids preferably interact with hard bases and soft acids preferably interact with soft bases. Hence, for a specific sulfocompound (e.g., DBT in this work), the softer the acidity of the adsorbent is, the stronger they will be interacted and accordingly the higher may be the adsorption capacity of the adsorbents. In addition, previous experiments have shown that the energy level of molecular orbitals (directly related to chemical hardness) for two-dimensional materials can be improved by doping other types of atoms6,32 (such as C, N, O, etc.). Here, we report that the adsorption capacity can be enhanced from h-BN to BCN by tuning the chemical hardness via doping C element into h-BN.

Table 1. Values of Different Descriptors of CDFT (Unit: au): εHOMO, εLUMO, Hardness (η), Electronegativity (χ), and HOMO−LUMO Gap species

εHOMO (au)

εLUMO (au)

η (au)

χ (au)

gap (eV)

BN model_1C model_2C model_3C model_4C model_5C

−0.30435 −0.28244 −0.24967 −0.24770 −0.24692 −0.24987

0.00585 0.00423 −0.15422 −0.15846 −0.17602 −0.17588

0.31020 0.28667 0.09545 0.08924 0.07090 0.07399

0.14925 0.13911 0.20195 0.20308 0.21147 0.21288

8.44 7.80 2.60 2.43 1.93 2.01

related to the donation ability, whereas LUMO is related to the back-donation ability.28 From pure h-BN to model_5C, it can be seen that the energy level of HOMO is elevated and the energy level of LUMO is declined, which indicate a higher donation and back-donation ability, thereby leading to a reduction of the band gap for these materials. Finally, it can be observed that the hardness of the BCN is reduced. In addition, the electronegativity was involved to confirm the acidity of the adsorbents. All of the obtained values (χ > 0.10290 au or 2.8 eV) show that these adsorbents can be treated as Lewis acids.28,29 On the basis of the above discussion, we know that lower hardness means higher adsorption strength. Hence, to enhance the adsorption capacity, the carbon-doping strategy may be feasible from the insight of DFT calculations and experimental facts. 2.2. Structure and Morphology of BCN. Motivated by the above theoretical insights and experimental facts, a series of BCN-x (x = 10−50, x is the amount of carbon resource vs boron acid) materials have been synthesized . First, to characterize the structure of the BCN, X-ray diffraction (XRD) and Fourier transform infrared (FTIR) spectra analyses were carried out. Figure 2A reveals a typical hexagonal structure of BCN with two

2. RESULTS AND DISCUSSION 2.1. Chemical Hardness and Frontier Orbitals of BCN Materials. In this work, we first analyzed the frontier orbitals and the chemical hardness for different C-content model BCN materials (model_xC, x = 1−5) by conceptual density functional theory (CDFT) (Figure S1, Supporting Information). In the model materials, the C content increases with increasing x value, which is used to simulate the trend of C-doping in real materials. Computational results show that frontier orbitals (highest occupied molecular orbital (HOMO) and lowest unoccupied molecular orbital (LUMO)) are composed of π or anti-π orbitals for both pure h-BN and model_xC (Figure 1). It is found that the 5386

DOI: 10.1021/acsomega.7b00795 ACS Omega 2017, 2, 5385−5394

ACS Omega

Article

Figure 2. (A, C) XRD pattern and (B) FTIR spectra of BN (a), BCN-10 (b), BCN-20 (c), BCN-30 (d), BCN-40 (e), and BCN-50 (f). (D) Distance of arrows corresponding to the (100) index of h-BN materials.

Figure 3. (a) UV−vis diffuse reflectance spectra of the samples; (b) band gap determination of the BCN-x samples from the (F(R)·E)2 vs E plots; (c) UV−vis spectra by TD-DFT calculation; (d) hardness values of the samples by DFT calculation.

broadened (002) and weak (100) peaks at 2θ = 25.2 and 42.6° (JCPDS card no. 01-073-2095).6,18 The corresponding interlayer distance of BCN-50 at 2θ = 25.2° is d = 3.51 Å, which is larger than the distance in boron nitride nanosheets (BNNSs).33 The (100) peak, corresponding to the distance between two adjacent

rings (Figure 2D), becomes gradually less intense and shifted to wide angle with increasing amount of carbon resource (Figure 2C). The corresponding distance for commercial BN is 2.56 Å, whereas for BN and BCN-50, it is reduced to 2.12 Å, which indicates disturbance of graphitic structures, potentially by the C5387

DOI: 10.1021/acsomega.7b00795 ACS Omega 2017, 2, 5385−5394

ACS Omega

Article

Figure 4. (a−c) Corresponding mappings of B, C, and N in BCN-50. (d) EEL spectrum of BCN-50.

doping effects.34 The FTIR spectra of BCN are much similar to those of h-BN, as shown in Figure 2B. The absorption peaks at approximately 810 cm−1 belong to B−N−B out-of-plane bending vibration, whereas the broad absorption bands at about 1375 cm−1 are ascribed to the B−N in-plane stretching vibration,35,36 which indicates that the BCN still maintains the hexagonal structure of h-BN. The absorption peaks of C−B or C−N bonds could not be found in the BCN-x, which may be ascribed to the typical overlap with B−N bands.6 Nonetheless, shifts of these two peaks were observed comparing with h-BN, which indicates a little change of chemical bonding environment around the bands, possibly influenced by the change of interlayer interactions. To prove that the carbon was indeed doped into the h-BN graphitic network in BCN materials, ultraviolet−visible diffuse reflectance spectroscopy (UV−vis DRS) was employed to analyze the excited state and band gap. As shown in Figure 3a, both the absorption spectra of pristine BN and BCN-x have the same absorption peak at 245 nm, attributing to the σ bond electron excitations of h-BN.37 It is noteworthy that there are distinct absorption peaks at 328 nm of BCN-x.6 Furthermore, the tail peaks in the visible region become stronger and stronger as the amount of carbon resource increases in BCN-x. These phenomena are on account of the carbon doping in the h-BN graphitic network and will be further confirmed by the following time-dependent DFT (TD-DFT) calculations. Similar results can be observed in the former papers.38−40 Hence, above results show that the introduction of carbon could change the electron excitation and the HOMO−LUMO status. This change can be also interpreted by color in macroscale that the colors of BCN-x become darker and darker (Figure S2). Furthermore, the change of HOMO−LUMO can directly affect the band gap or other physicochemical properties (e.g., hardness). It can be seen in Figure 3b that the indirect band gap values of BCN-x decrease

gradually, thereby leading to a decrease of the hardness of the related materials. To reach a deep understanding of the new peak at 328 nm for BCN-x, UV−vis spectra of BN and model_xC were calculated by TD-DFT, as shown in Figure 3c. Obvious red shift to high wavelength of the main excitation peaks can be observed for these three BCN-xC model materials. Moreover, strong absorption in the visible region from theoretical computation also gives another evidence to confirm the Cdoping effects for h-BN. Above observations result in a reduced HOMO−LUMO gap by doping C atoms into h-BN (Table 1). The reduced HOMO−LUMO gap allows the excitation of electrons in a low energy range, thus leading to a red shift in the UV−vis spectrum. Therefore, corresponding hardness of BCN-x could also follow the same trend, which gives the direct evidence to prove that the introduction of carbon can reduce the hardness of BN, from UV−vis spectra. The calculated results of hardness are shown in Figure 3d, which agree well with the experimental results. Element mapping of BCN-50 (Figure 4a−c) gave a uniform distribution of B, C, and N throughout the whole selected area, proving homogeneity of the ternary BCN material. To further confirm the chemical environment of all elements, we carried out electron energy loss spectroscopy (EELS) measurements on the K-edge absorption for B, C, and N. The EEL spectrum of the BCN-50 in Figure 4d clearly shows the K-shell ionization edges at about 190, 284, and 401 eV for B, C, and N, respectively. The two peaks at each core-edge fine structure is the characteristic of the sp2-hybridization state, corresponding to π* and θ* bands, respectively.41−43 Thereby it indicates that B, C, and N are in a commonly sp2-hybridized, 2D-conjugated electron system, rather than a physical mixture. Above experimental as well as theoretical discussions give direct evidences that the HOMO− LUMO gap (or chemical hardness) can be tuned by forming 5388

DOI: 10.1021/acsomega.7b00795 ACS Omega 2017, 2, 5385−5394

ACS Omega

Article

Figure 5. (a) SEM of BN, (b) SEM of BCN-40, (c) TEM of BN, and (d) TEM of BCN-40.

Figure 6. (a) XPS survey spectra of BN and BCN-50, (b) C 1s of BN, (c) C 1s of BCN-50, and (d) atomic concentrations of B, C, and N in BN and BCN-50.

5389

DOI: 10.1021/acsomega.7b00795 ACS Omega 2017, 2, 5385−5394

ACS Omega

Article

Figure 7. (a) Effect of different adsorbents on DBT adsorption. Experimental conditions: m (adsorbent) = 0.05 g; V (oil) = 20 mL; T = 298 K; Cs (oil) = 300 ppm. (b) Stable configurations and adsorption energies of DBT by BN and model_5C using DFT calculation. (c) Effect of DBT with different sulfur initial concentrations on DBT adsorption by BCN-50. Experimental conditions: m (adsorbent) = 0.05 g; V (oil) = 20 mL; T = 298 K. (d) Effect of temperature by BCN-50. Experimental conditions: m (adsorbent) = 0.05 g; V (model oil) = 20 mL; Cs (oil) = 300 ppm.

different BCN-x samples and thus may finally change the adsorption capacity of BCN-x (see Section 2.4). For the morphology, the transmission electron microscopy (TEM) and scanning electron microscopy (SEM) images of the BCN-x are depicted in Figure 5, which gives significant information of morphology and microstructure. From the SEM images, both BN and BCN-x keep the nanosheet structure. The nanosheets of BCN-x are fragmented and show the “water plant” structure with increasing amount of carbon resource (Figures S3 and S4). This observation has been further testified by TEM. It can be seen in BCN-10 (Figure S4a) that there are some holes with different pore size contributing to the fragmentation of nanosheets. These holes will reduce the periodic boundary of the graphitic rings and thus lead to a reduced intensity of (100) plane, as depicted in Figure 2. Then, the nanosheets obviously and regularly fragment into lathy water plant structure in BCN40 (Figure 5b). Before BCN-40, the cetyltrimethylammonium chloride (CTAC) does not act as template until the concentration is up to critical micelle concentration. 2.3. Carbon Content of BCN. X-ray photoelectron spectroscopy (XPS) was used to further investigate the chemical composition of BCN comparing with h-BN. As shown in Figure 6a, both BN and BCN have B, N, C, and O elements. The peak of C 1s in BCN-50 is visibly stronger than BN because the content of C in BCN is increased. By calculating from the XPS survey spectrum, the atomic concentrations of B, C, N, and O in BN and BCN-50 are listed in Figure 6d. It can be seen that the contents of both boron and nitride in BCN-50 decrease comparing with BN, whereas their carbon content increases from 12.4 to 23.3%, which reaches close to an ideal stoichiometry of BCN. In addition, other BCN samples show the same trends for atomic concentrations (Table S1). Elemental analysis of other BCN-x samples (Figure S5) revealed that the carbon content is gradually increased with increasing amounts of CTAC. The BN also

contains 12.4% carbon due to the synthetic method without removing the adventitious graphitic carbon on the surface and the adsorbed carbon dioxide. From the deconvolution of the C 1s spectrum of BN in Figure 6b, evidence can be given that there are two subpeaks centered at 284.7 and 288.3 eV, which correspond to C−C (graphitic carbon) and C−O (CO2) components, respectively.18,42 The corresponding C 1s spectrum of BCN-50 shown in Figure 6c can be deconvoluted into four subpeaks. In addition to the two same subpeaks in BN, the binding energy of 285.5 eV was ascribed to the C−N bond and that of 283.7 eV was assigned to the C−B bond,18,42,44,45 which suggest that the carbon atoms indeed substitute partial B and N atoms. Apart from these, it is interesting to find that the percentage of C bonded to B is higher than that of C bonded to N species. Accordingly, above results show that the content of nitride decreases more quick compared with the amount of boron. To give an insight into this phenomenon, a substitution model based on DFT was constructed (see Section 4.1). The substitution energy was employed to judge whether B or N is easier to be substituted by C atom. The calculated results show that the substitution energy of B is 40.2 kcal mol−1, whereas that of N is 80.3 kcal mol−1. Although the substitution reactions of both B and N are endothermic, the substitution energy of N is much higher than that of B. Hence, we suggest that the N atoms in hBN are easier to be substituted by C atoms. In this study, the highest carbon content of BCN reaches to 23.3%, which is similar to previous C-doping works.6,41,42,46,47 2.4. Adsorption Performance of BCN. In the following, to verify the performance of adsorption capacity by tuning the chemical hardness, the removal of DBT in model oil by adsorption was tested. The data of adsorption experiment are shown in Section 4.2. The corresponding adsorption capacities of BCN-x are depicted in Figure 7a. It can be seen that the BCN materials show higher adsorption capacity than pristine h-BN. 5390

DOI: 10.1021/acsomega.7b00795 ACS Omega 2017, 2, 5385−5394

ACS Omega

Article

the distance between the plane of DBT and BN (or BCN_5C) is about 3.22 Å (or 3.24 Å), which is a typical length for π−π interactions. The calculated adsorption energy for the BN_DBT system is −20.12 kcal mol−1, which clearly shows a strong interaction between BN and DBT. The adsorption energy of BCN_5C_DBT is up to −75.70 kcal mol−1, which is much higher than that of BN and thus shows an enhancing adsorption capacity by C-doping effect. This stronger interaction has been reported in previous works, named pancake bond.48 In addition, other model_xC_DBT systems show the same trend of the adsorption energies (Table S4). These DFT results give direct evidence that the adsorption capacity of h-BN can be enhanced by C-doping.

With respect to the BCN-x sample, the adsorption performance is enhanced when the amount of carbon resource increased from 10 to 50% (i.e., carbon content: 12.4−23.3%, Table S1). BCN-50 has the highest adsorption capacity (27.43 mg g−1), which is much higher than the pristine h-BN and commercial BN (Figure 7a). In the former work, to demonstrate the synergistic effect of C-doping and the porous structure of BN, the individual Cdoping was also investigated using glucose as the carbon source instead of IL. It can be found that the C-doping leads to some improvement in adsorption.38 In addition, it now seems that when more carbon atoms are doped into the graphitic network of h-BN, the materials will show higher adsorption capacity for DBT. The higher adsorption capacity of the current BN sample compared to commercial BN may be ascribed to the higher concentration of C, which also shows that the carbon can enhance the adsorption capacity. In general, adsorbents with larger surface areas can lead to higher adsorption performance. This is usually correct because the materials with larger surface areas will expose more adsorption sites to interact with adsorbate. On the contrary, the BCN-x sample shows lesser surface area (BCN-50: 160 m2 g−1) than pristine h-BN (772 m2 g−1) in this work, as measured by Brunauer−Emmett−Teller (BET) analysis (Figure S6 and Table S2), which may be due to the effect of carbon introduction. Similar results can be seen in the former papers. It indicates that the surface area is not the dominant effect for adsorption capacity in the current systems. However, we found that the hardness can explain the structure activity relationship. As depicted in Figure 3d, the values of hardness for model_xC decrease with increasing amount of carbon atoms. Accordingly, it is quite obvious that the lower hardness the adsorbent exhibits, the higher adsorption capacity the adsorbent shows. For an adsorbent with Lewis acidity, the approach to reduce its hardness is beneficial to enhance its adsorption capacity for aromatic sulfocompounds with Lewis basicity (e.g., DBT). In terms of HSAB theory, the Lewis acid (BCN) can interact with the Lewis base (DBT) to form a Lewis acid−base complex. The soft acid can have strong complexation with the soft base, especially. It is difficult to change the property of DBT in oil phase, but changing the property of adsorbent is possible. Doping carbon atoms into the graphitic network of h-BN can reduce the whole hardness, which is useful to enhance the adsorption capacity of DBT. To investigate the effect of temperature, the adsorption of DBT on BCN-50 was studied at different temperatures (298, 318, and 338 K). The results are presented in Figure 7d and show that the adsorption capacity decreased with increasing temperature. It is proved that the adsorption process is exothermic. Thus, 298 K is in favor of adsorptive desulfurization. The adsorptive desulfurization of BCN-x under various initial sulfur concentrations was also conducted and presented in Figure 7c. The adsorption capacity increases with increasing initial sulfur concentration, which shows that more efficient adsorption of DBT on BCN samples can be obtained at high concentration. In addition, reaction kinetic analysis shows that it fits well with the pseudo-second-order kinetics (Figure S7 and Table S3), which is consistent with previous works in similar systems.8 2.5. Adsorption Mechanism. Although the HSAB theory can be used to explain the activity of adsorption, the essence of adsorption between adsorbate and adsorbent molecule cannot be described iconically and directly. Therefore, adsorption models of BN_DBT and model_xC_DBT were built and the adsorption energies were calculated by DFT method to further analyze the difference between BN and BCN. It can be seen in Figure 7b that

3. CONCLUSIONS In brief, an efficient strategy can be used to enhance the adsorption capacity by introducing C element into h-BN materials. First, frontier orbitals, electronegativity, chemical hardness, and energy gap were examined by CDFT. Results show that the HOMO−LUMO gap and the chemical hardness are reduced after the C atoms are doped into the graphitic network of h-BN. XRD, FTIR, UV−vis DRS, XPS, and EELS technologies show that the C element is successfully doped into the BCN-x materials. The as-prepared BCN-x materials show better adsorption capacity (e.g., 27.43 mg g−1 for BCN-50) than the commercial h-BN. It is found that the adsorption capacity correlates well with the C content in the BCN-x, which is consistent with the results predicted by theoretical calculation. The enhancement of the adsorption capacity is ascribed to the reduction of the HOMO−LUMO gap and the chemical hardness after C-doping, which was confirmed by UV−vis spectra and theoretical calculations. Hence, we propose that the enhancement of adsorption capacity can be correlated to the reducing of the chemical hardness, from h-BN to BCN by C-doping strategy. This strategy may be expanded to other similar adsorptive systems, where the π−π interaction is the dominant driven force. 4. METHODS 4.1. Computational Details. For adsorption systems, there are various types of noncovalent interactions, such as induction, electrostatic, and dispersion interactions. Traditional hybrid functionals, like B3LYP, can give reasonable results for the systems that mainly consist of induction and electrostatic interactions. However, they are not good enough for the systems in which dispersion interaction is dominant. Hence, the Minnesota functional (M06-2X) was implemented in this work because of its good performance on noncovalent systems. In addition, 6-311+G(2d) basis set was used for all of the calculations. Geometry optimization and energy calculations were all performed using the Gaussian 09 suite of programs. For the model of h-BN, a cluster model, which contains 52 atoms (B21N21H16), was employed. This model contains 12 aromatic rings, which should be big enough to study the adsorption of DBT (three aromatic rings). In addition, similar models have been widely used to model properties of h-BN. The C-doped hBN models were constructed by gradually dispersing C atoms into the h-BN network, resulting in five cluster models, such as B21CN20H16, B21C2N19H16, B21C3N18H16, B21C4N17H16, and B21C5N16H16. For the stabilities of these BCN materials, frequency calculations were performed and no imaginary frequencies were found, which means all of the related model materials are stable. 5391

DOI: 10.1021/acsomega.7b00795 ACS Omega 2017, 2, 5385−5394

ACS Omega

Article

4.1.1. Substitution Energy. To consider the doping effect on h-BN, a substitution model was constructed by substituting boron/nitride atom with carbon atom. The former study shows that B, C, and N can substitute with each other and form various new structures due to their similar atomic parameters. However, which atom will be substituted, the B atom or N atom? This question is meaningful for understanding the formation mechanism of BCN. Herein, the concept of substitution energy was adopted to quantitatively analyze which atom will be substituted by the C atoms. Equation 1 shows the definition of the substitution energy (Es) Es = (E BCN + Esub) − (E BN + EC)

solvent evaporated completely. The obtained white solid was ground a little. Then, the powder was calcined at 1000 °C for 2 h in N2 atmosphere in a tube furnace. The obtained products were marked as C/BN. Moreover, the boron nitride nanosheets were similarly prepared without CTAC and marked as BNNSs.6 4.2.3. Measurements. Fourier transform infrared spectra were recorded on a Nicolet Model Nexus 470 FTIR instrument using KBr disk at room temperature. XRD spectra were recorded on a Bruker D8 diffractometer with high-intensity Cu Kα (λ = 1.54 Å). Field-emission scanning electron microscopy (SEM) measurements were carried out with a field-emission scanning electron microscope (JEOL JSM-7001F). Transmission electron microscopy (TEM) images were recorded on a JEOL-JEM-2010 microscope (JEOL, Japan) operating at 200 kV. EELS spectrum was recorded on EDAX FP9761/70 by FEI TECNAI G2 F20 STWIN. UV−vis spectra were recorded on an UV-2450 spectrophotometer (Shimadzu Corporation, Japan). X-ray photoelectron spectroscopy (XPS) was performed on a VG MultiLab 2000 spectrometer using Mg KR (1253.6 eV) radiation. BET surface area and pore volume were obtained from 77 K N2 adsorption−desorption isotherms using a TriStar II 3020 (Micromeritics Instrument Corporation). 4.2.4. Adsorption Process. The model oil used for experiments was prepared by dissolving DBT in n-octane, and the concentrations of sulfur were 100, 300, 500, and 800 ppm. In a typical adsorptive desulfurization experiment, 0.05 g of adsorbent was added to 20 mL of model oil in a 50 mL conical flask. Then, the flask was fixed in a shaker with a water bath maintained at a temperature of 298 K and agitated for 120 min. 4.2.5. Analysis Methods. The residual sulfur concentration in model oil after the adsorption was analyzed by a gas chromatography-flame ionization detector (GC-FID) with tetradecane as the internal standard (Agilent 7890A; HP-5, 30 m × 0.32 mm i.d. × 0.25 mm; FID: Agilent). The adsorbed amount (qt) of sulfur compounds was calculated by the following equation

(1)

where EBCN and EBN are the total energies of BN with and without substitution, respectively, Esub is the energy of the substituted atom (boron or nitride), EC is the energy of carbon atom. The energies of the substituted atom (boron, carbon, or nitride) were calculated in the single atomic state. The electronic configurations are B (2s22p1), C(2s22p2), and N(2s22p3), which correspond to spin multiplicities of 2, 3, and 4, respectively. Their calculated energies are −24.6485, −37.8406, and −54.5846 au, respectively. 4.1.2. Conceptual Density Functional Theory (CDFT). CDFT is using DFT descriptors to provide qualitative insights and to quantify intuitive chemical concepts.49,50 Commonly used concepts in CDFT are ionization energy (I), electron affinity (A), hardness (η), electronegativity (χ), and so on. Hardness means the resistance of the chemical potential to change in the number of electrons. In Table 1, the values of ionization energy and electron affinity are omitted because they equal −εHOMO and −εLUMO. The corresponding operational definition of hardness is the finite difference formula

η=I−A or approximation through Koopman’s theorem η = εLUMO − εHOMO

(2)

(3)

qt =

showing hardness is also due to the HOMO−LUMO gap. This gap can describe the ability of electron transition. Generally speaking, if the HOMO−LUMO gap is less than 3.0 eV, the material would have absorption in the visible region. Electronegativity is defined as follows

where C0 and Ct (mg L−1) are the initial sulfur concentration and the sulfur concentration at time t, respectively, V (L) is the volume of the oil, and m (g) is the mass of the adsorbent.



⎛ δE ⎞ ⎛ ∂E ⎞ χ = −⎜ ⎟ = −⎜ ⎟ ⎝ ∂N ⎠v ⎝ δρ(r ) ⎠v

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acsomega.7b00795. Structures of model materials (Figure S1), colors for different BCN-x samples (Figure S2), SEM and TEM images of BCN-x (Figures S3 and S4), atomic concentrations in BCN-x (Table S1), XPS spectra (Figure S5), N2 adsorption−desorption isotherms (Figure S6), textual properties of different samples (Table S2), kinetic model and parameters for DBT adsorption (Figure S7, Table S3), adsorptive energies for DBT (Table S4) (PDF)

where E is the ground-state energy, ρ(r) is the electronic density, N is the total electron number. There exist practical routes to estimate chemical potential in quantitative DFT calculations. According to Mulliken χ=

(C 0 − C t )V m

(I + A ) 2

4.2. Experimental Details. 4.2.1. Materials. Boric acid, urea, cetyltrimethylammonium chloride (CTAC), and n-octane were purchased from Sinopharm Chemical Reagent Co. Ltd. DBT was purchased from Sigma-Aldrich. 4.2.2. Synthesis of Adsorbents. In a typical synthetic procedure, carbon-doped boron nitride nanosheets were synthesized. Boric acid and urea with the molar ratio of 1:4 were dissolved in 20 mL of methyl alcohol. Then, cetyltrimethylammonium chloride as carbon source was added into the solution and stirred. The mixture was heated at 60 °C until the



AUTHOR INFORMATION

Corresponding Authors

*E-mail: [email protected] (W. S. Zhu). *E-mail: [email protected] (H. M. Li). ORCID

Huaming Li: 0000-0002-9538-5358 5392

DOI: 10.1021/acsomega.7b00795 ACS Omega 2017, 2, 5385−5394

ACS Omega

Article

Notes

and Ternary Atomic Layers Built from Carbon, Boron, and Nitrogen. Adv. Mater. 2012, 24, 4878−4895. (18) Karbhal, I.; Devarapalli, R. R.; Debgupta, J.; Pillai, V. K.; Ajayan, P. M.; Shelke, M. V. Facile Green Synthesis of BCN Nanosheets as HighPerformance Electrode Material for Electrochemical Energy Storage. Chem. − Eur. J. 2016, 22, 7134−7140. (19) Babich, I. V.; Moulijn, J. A. Science and Technology of Novel Processes for Deep Desulfurization of Oil Refinery Streams: A Review. Fuel 2003, 82, 607−631. (20) Kulkarni, P. S.; Afonso, C. A. M. Deep Desulfurization of Diesel Fuel Using Ionic Liquids: Current Status and Future Challenges. Green Chem. 2010, 12, 1139−1149. (21) Zhang, Z. Y.; Shi, T. B.; Jia, C. Z.; Ji, W. J.; Chen, Y.; He, M. Y. Adsorptive Removal of Aromatic Organosulfur Compounds over the Modified Na-Y Zeolites. Appl. Catal., B 2008, 82, 1−10. (22) Hernandez-Maldonado, A. J.; Yang, R. T. Desulfurization of Commercial Liquid Fuels by Selective Adsorption Via Pi-Complexation with Cu(I)-Y Zeolite. Ind. Eng. Chem. Res. 2003, 42, 3103−3110. (23) Hernández-Maldonado, A. J.; Yang, R. T. Desulfurization of Liquid Fuels by Adsorption Via Π Complexation with Cu(I)-Y and Ag-Y Zeolites. Ind. Eng. Chem. Res. 2003, 42, 123−129. (24) Chen, H.; Wang, Y. H.; Yang, F. H.; Yang, R. T. Desulfurization of High-Sulfur Jet Fuel by Mesoporous Π Complexation Adsorbents. Chem. Eng. Sci. 2009, 64, 5240−5246. (25) Gao, J. J.; Li, H. Q.; Zhang, H. X.; Lu, Y. Z.; Meng, H.; Li, C. X. Removal Mechanism of Thiophenic Compounds in Model Oil by Inorganic Lewis Acids. Ind. Eng. Chem. Res. 2012, 51, 4682−4691. (26) Li, H. P.; Chang, Y. H.; Zhu, W. S.; Zhu, S. W.; Jiang, W.; Zhang, M.; Zhou, Y. W.; Xia, J. X.; Li, H. M. The Selectivity for Sulfur Removal from Oils: An Insight from Conceptual Density Functional Theory. AIChE J. 2016, 62, 2087−2100. (27) Pal, S.; Vivekchand, S. R. C.; Govindaraj, A.; Rao, C. N. R. Functionalization and Solubilization of BN Nanotubes by Interaction with Lewis Bases. J. Mater. Chem. 2007, 17, 450−452. (28) Li, H.; Zhu, W.; Chang, Y.; Jiang, W.; Zhang, M.; Yin, S.; Xia, J.; Li, H. Theoretical Investigation of the Interaction between Aromatic Sulfur Compounds and [BMIM]+[FeCl4]− Ionic Liquid in Desulfurization: A Novel Charge Transfer Mechanism. J. Mol. Graphics Modell. 2015, 59, 40−49. (29) Yu, M.; Li, Z.; Xia, Q.; Xi, H.; Wang, S. Desorption Activation Energy of Dibenzothiophene on the Activated Carbons Modified by Different Metal Salt Solutions. Chem. Eng. J. 2007, 132, 233−239. (30) Xiao, J.; Li, Z.; Liu, B.; Xia, Q. B.; Yu, M. X. Adsorption of Benzothiophene and Dibenzothiophene on Ion-Impregnated Activated Carbons and Ion-Exchanged Y Zeolites. Energy Fuels 2008, 22, 3858− 3863. (31) Xiao, J.; Bian, G. A.; Zhang, W.; Li, Z. Adsorption of Dibenzothiophene on Ag/Cu/Fe-Supported Activated Carbons Prepared by Ultrasonic-Assisted Impregnation. J. Chem. Eng. Data 2010, 55, 5818−5823. (32) Zhang, J. S.; Chen, Y.; Wang, X. C. Two-Dimensional Covalent Carbon Nitride Nanosheets: Synthesis, Functionalization, and Applications. Energy Environ. Sci. 2015, 8, 3092−3108. (33) Zhu, W.; Gao, X.; Li, Q.; Li, H.; Chao, Y.; Li, M.; Mahurin, S. M.; Li, H.; Zhu, H.; Dai, S. Controlled Gas Exfoliation of Boron Nitride into Few-Layered Nanosheets. Angew. Chem., Int. Ed. 2016, 55, 10766− 10770. (34) Pakdel, A.; Bando, Y.; Golberg, D. Nano Boron Nitride Flatland. Chem. Soc. Rev. 2014, 43, 934−959. (35) Li, M.; Zhu, W.; Zhang, P.; Chao, Y.; He, Q.; Yang, B.; Li, H.; Borisevich, A.; Dai, S. Graphene-Analogues Boron Nitride Nanosheets Confining Ionic Liquids: A High-Performance Quasi-Liquid Solid Electrolyte. Small 2016, 12, 3535−3542. (36) Li, J.; Xiao, X.; Xu, X.; Lin, J.; Huang, Y.; Xue, Y.; Jin, P.; Zou, J.; Tang, C. Activated Boron Nitride as an Effective Adsorbent for Metal Ions and Organic Pollutants. Sci. Rep. 2013, 3, No. 3208. (37) Wu, P.; Zhu, W.; Dai, B.; Chao, Y.; Li, C.; Li, H.; Zhang, M.; Jiang, W.; Li, H. Copper Nanoparticles Advance Electron Mobility of

The authors declare no competing financial interest.



ACKNOWLEDGMENTS This work was financially supported by the National Natural Science Foundation of China (Nos. 21646001, 21506080, and 21406092) and A Project Funded by the Priority Academic Program Development of Jiangsu Higher Education Institutions.



REFERENCES

(1) Balendhran, S.; Walia, S.; Nili, H.; Sriram, S.; Bhaskaran, M. Elemental Analogues of Graphene: Silicene, Germanene, Stanene, and Phosphorene. Small 2015, 11, 640−652. (2) Fu, Q.; Bao, X. Surface Chemistry and Catalysis Confined under Two-Dimensional Materials. Chem. Soc. Rev. 2017, 46, 1842−1874. (3) Peng, Q.; Guo, J. X.; Zhang, Q. R.; Xiang, J. Y.; Liu, B. Z.; Zhou, A. G.; Liu, R. P.; Tian, Y. J. Unique Lead Adsorption Behavior of Activated Hydroxyl Group in Two-Dimensional Titanium Carbide. J. Am. Chem. Soc. 2014, 136, 4113−4116. (4) Sun, Y.; Gao, S.; Lei, F. C.; Xie, Y. Atomically-Thin TwoDimensional Sheets for Understanding Active Sites in Catalysis. Chem. Soc. Rev. 2015, 44, 623−636. (5) Zhang, H. Ultrathin Two-Dimensional Nanomaterials. ACS Nano 2015, 9, 9451−9469. (6) Huang, C.; Chen, C.; Zhang, M.; Lin, L.; Ye, X.; Lin, S.; Antonietti, M.; Wang, X. Carbon-Doped BN Nanosheets for Metal-Free Photoredox Catalysis. Nat. Commun. 2015, 6, No. 7698. (7) Shi, Y.; Hamsen, C.; Jia, X. T.; Kim, K. K.; Reina, A.; Hofmann, M.; Hsu, A. L.; Zhang, K.; Li, H. N.; Juang, Z. Y.; Dresselhaus, M. S.; Li, L. J.; Kong, J. Synthesis of Few-Layer Hexagonal Boron Nitride Thin Film by Chemical Vapor Deposition. Nano Lett. 2010, 10, 4134−4139. (8) Xiong, J.; Zhu, W. S.; Li, H. P.; Ding, W. J.; Chao, Y. H.; Wu, P. W.; Xun, S. H.; Zhang, M.; Li, H. M. Few-Layered Graphene-Like Boron Nitride Induced a Remarkable Adsorption Capacity for Dibenzothiophene in Fuels. Green Chem. 2015, 17, 1647−1656. (9) Zhang, S. J.; Lian, G.; Si, H. B.; Wang, J.; Zhang, X.; Wang, Q. L.; Cui, D. L. Ultrathin BN Nanosheets with Zigzag Edge: One-Step Chemical Synthesis, Applications in Wastewater Treatment and Preparation of Highly Thermal-Conductive BN-Polymer Composites. J. Mater. Chem. A 2013, 1, 5105−5112. (10) Lian, G.; Zhang, X.; Zhang, S. J.; Liu, D.; Cui, D. L.; Wang, Q. L. Controlled Fabrication of Ultrathin-Shell BN Hollow Spheres with Excellent Performance in Hydrogen Storage and Wastewater Treatment. Energy Environ. Sci. 2012, 5, 7072−7080. (11) Liu, D.; Lei, W.; Qin, S.; Chen, Y. Template-Free Synthesis of Functional 3d BN Architecture for Removal of Dyes from Water. Sci. Rep. 2014, 4, No. 4453. (12) Chao, Y.; Zhu, W.; Chen, J.; Wu, P.; Wu, X.; Li, H.; Han, C.; Yan, S. Development of Novel Graphene-Like Layered Hexagonal Boron Nitride for Adsorptive Removal of Antibiotic Gatifloxacin from Aqueous Solution. Green Chem. Lett. Rev. 2014, 7, 330−336. (13) Song, Y.; Li, B.; Yang, S.; Ding, G.; Zhang, C.; Xie, X. Ultralight Boron Nitride Aerogels Via Template-Assisted Chemical Vapor Deposition. Sci. Rep. 2015, 5, No. 10337. (14) Lei, W.; Portehault, D.; Liu, D.; Qin, S.; Chen, Y. Porous Boron Nitride Nanosheets for Effective Water Cleaning. Nat. Commun. 2013, 4, No. 1777. (15) Xiong, J.; Yang, L.; Chao, Y. H.; Pang, J. Y.; Wu, P. W.; Zhang, M.; Zhu, W. S.; Li, H. M. A Large Number of Low Coordinated Atoms in Boron Nitride for Outstanding Adsorptive Desulfurization Performance. Green Chem. 2016, 18, 3040−3047. (16) Xiong, J.; Yang, L.; Chao, Y. H.; Pang, J. Y.; Zhang, M.; Zhu, W. S.; Li, H. M. Boron Nitride Mesoporous Nanowires with Doped Oxygen Atoms for the Remarkable Adsorption Desulfurization Performance from Fuels. ACS Sustainable Chem. Eng. 2016, 4, 4457−4464. (17) Song, L.; Liu, Z.; Reddy, A. L. M.; Narayanan, N. T.; TahaTijerina, J.; Peng, J.; Gao, G. H.; Lou, J.; Vajtai, R.; Ajayan, P. M. Binary 5393

DOI: 10.1021/acsomega.7b00795 ACS Omega 2017, 2, 5385−5394

ACS Omega

Article

Graphene-Like Boron Nitride for Enhanced Aerobic Oxidative Desulfurization. Chem. Eng. J. 2016, 301, 123−131. (38) Xiong, J.; Zhu, W.; Li, H.; Yang, L.; Chao, Y.; Wu, P.; Xun, S.; Jiang, W.; Zhang, M.; Li, H. Carbon-Doped Porous Boron Nitride: Metal-Free Adsorbents for Sulfur Removal from Fuels. J. Mater. Chem. A 2015, 3, 12738−12747. (39) Chen, X.; Burda, C. The Electronic Origin of the Visible-Light Absorption Properties of C-, N- and S-Doped TiO2 Nanomaterials. J. Am. Chem. Soc. 2008, 130, 5018−5019. (40) Zhang, Y.; Mori, T.; Ye, J.; Antonietti, M. Phosphorus-Doped Carbon Nitride Solid: Enhanced Electrical Conductivity and Photocurrent Generation. J. Am. Chem. Soc. 2010, 132, 6294−6295. (41) Iyyamperumal, E.; Wang, S.; Dai, L. Vertically Aligned BCN Nanotubes with High Capacitance. ACS Nano 2012, 6, 5259−5265. (42) Raidongia, K.; Nag, A.; Hembram, K. P.; Waghmare, U. V.; Datta, R.; Rao, C. N. BCN: A Graphene Analogue with Remarkable Adsorptive Properties. Chem. − Eur. J. 2010, 16, 149−157. (43) Song, L.; Liu, Z.; Reddy, A. L.; Narayanan, N. T.; Taha-Tijerina, J.; Peng, J.; Gao, G.; Lou, J.; Vajtai, R.; Ajayan, P. M. Binary and Ternary Atomic Layers Built from Carbon, Boron, and Nitrogen. Adv. Mater. 2012, 24, 4878−4895. (44) Paul, R.; Voevodin, A. A.; Zemlyanov, D.; Roy, A. K.; Fisher, T. S. Microwave-Assisted Surface Synthesis of a Boron-Carbon-Nitrogen Foam and Its Desorption Enthalpy. Adv. Funct. Mater. 2012, 22, 3682− 3690. (45) Lei, W.; Portehault, D.; Dimova, R.; Antonietti, M. Boron Carbon Nitride Nanostructures from Salt Melts: Tunable Water-Soluble Phosphors. J. Am. Chem. Soc. 2011, 133, 7121−7127. (46) Kang, Y.; Chu, Z.; Zhang, D.; Li, G.; Jiang, Z.; Cheng, H.; Li, X. Incorporate Boron and Nitrogen into Graphene to Make BCN Hybrid Nanosheets with Enhanced Microwave Absorbing Properties. Carbon 2013, 61, 200−208. (47) Ci, L.; Song, L.; Jin, C. H.; Jariwala, D.; Wu, D. X.; Li, Y. J.; Srivastava, A.; Wang, Z. F.; Storr, K.; Balicas, L.; Liu, F.; Ajayan, P. M. Atomic Layers of Hybridized Boron Nitride and Graphene Domains. Nat. Mater. 2010, 9, 430−435. (48) Tian, Y. H.; Sumpter, B. G.; Du, S.; Huang, J. Pancake Pi-Pi Bonding Goes Double: Unexpected 4e/All-Sites Bonding in Boron- and Nitrogen-Doped Phenalenyls. J. Phys. Chem. Lett. 2015, 6, 2318−2325. (49) Liu, S. B. Conceptual Density Functional Theory and Some Recent Developments. Acta Phys.-Chim. Sin. 2009, 25, 590−600. (50) Geerlings, P.; De Proft, F.; Langenaeker, W. Conceptual Density Functional Theory. Chem. Rev. 2003, 103, 1793−1873.

5394

DOI: 10.1021/acsomega.7b00795 ACS Omega 2017, 2, 5385−5394