Can boron nitride analog of carbon nanoneedle exist?

1 Introduction

Carbon is the “heart” of benzene which, along with its innumerous derivatives, forms the basis oforganic chemistry. Carbon also forms a variety of nanostructures such as fullerenes, nanotubes, nanowires etc. which are useful for many technological applications [1, 2]. Borazine is an inorganic analogue of the benzene, popularly known as the inorganic benzene [3]. It can be realized by replacing a pair of C atoms by a boron nitride (BN) unit; it was synthesized by reaction of diborane and ammonia in 1926 [4]. Borazine possesses a number of features analogous to the benzene, however, the polarity of the BN bond causes borazine to show a reactivity pattern different from that of benzene. For instance, borazine prefers addition reactions unlike benzene whose chemistry is based on substitution reactions. The BN analogues of many carbon nanostructures such as boron-nitride nanotubes, boron-nitride nanosheets etc. have also been studied. Many studies have been carried out on the understanding of analogy BN bond to CC bond [5, 6] in various heterocyclic systems [7–9].

The nanotubes find important applications in drug delivery systems. Being extremely thin but very long, they offer a large surface area on which the required drug can be bound. It has been demonstrated that the chemically modified carbon nanotubes can act as nanoneedles which can easily pass through biological barriers and penetrate a variety of cell types [10–12]. This finding reveals the potential of carbon nanoneedles (CNNs) as a new form of direct drug delivery [13]. The electronic properties of ultrathin carbon needle-like and tube-like nanostructures, which are tighter than the smallest single wall nanotubes, have been explored [14]. In a subsequent study [15], a series of (m×n) CNNs have been modelled and optimized, where m and n are the number of C atoms in a single layer and the number of layers in the CNN considered, respectively for m = 4, 6, 8 and n = 1–10, 15, 20. It was established that all CNNs are stable irrespective of their dimensions and hence, can be realized synthetically. In this communication, we address the question; can BN analogues of CNN exist?

2 Computational methods

In order to check the possibility of BN nanoneedles (BNNNs), initially we have modelled (m×n) CNNs for m = 4 and n = 1–5. For instance, we have shown (4×4) CNN [15] and its hypothetical BN analog in Fig. 1. These hypothetical BNNNs have been fully optimized without any symmetry constraint using density functional theory (DFT) at B3LYP/6-31 G(d) level. The present computational scheme has already been used in a previous study on CNN [15] and B3LYP [16, 17] method has also been used in our recent study [18]. Vibrational frequency calculations and natural population analysis (NPA) [19] have been carried out within same computational scheme. In a recent study, we have checked the performance of NPA charges as compared to various other population schemes [20]. All calculations have been performed via Gaussian 09 program package [21].

3 Results and discussion

After optimization, we find that the equilibrium structures become distorted from hypothetical BNNN structures. The optimized structures along with corresponding vibrational infrared (IR) spectra are shown in Figs. 2 and 3. One can note that BN analogs of CNN are no longer needle like structures. This is due to the electronegativity difference between B and N which creates a polarity in BN bond, unlike the covalent CC bond. In order to further explore this fact, we analyze NBO charges on atoms in (4×1) and (4×2) BN and CNN structures in Fig. 4. It is apparent that (4×1) and (4×2) CNN structures become planar (C_2h) and octahedral (O_h) in which each C possesses –0.23 and –0.24 e, respectively due to charge transfer from terminal H atoms. On the contrary, BN-structures become distorted to C_2v and C_3v structures. For instance, in (4×1) C_2v structure, the dihedral BNBN becomes 155. This distortion results due to the repulsion created by non-bonding electrons (lone pairs) of N atoms as well as polarity of BN bond. One can see that B and N atoms carry +0.76 e or +0.73 e and –1.11 e or –1.12 e charges in (4×1) or (4×2) structures (see Fig. 4). Note that there is a charge transfer to and from H atoms attached to B and N atoms, respectively due to difference in their electronegativity.

It is not possible to realize the BNNNs i.e. BN analogs of CNN, however, the optimized BN-structures (given in Figs. 2, 3) belong to at least some local minima in the potential energy surfaces. This is confirmed by vibrational frequency calculations which provide all real and positive values for the optimized structures. The bond-lengths (B–N) of the optimized structures are collected in Table 1. The average bond-length increases with the increase in the number of BN-layer (n). This fact is similar to that observed in the case of CNNs [15]. In Table 1, the binding energies per atom (ΔE) of BN-structures are also listed which are calculated as below: Δ E = ( x E [ H ] + y E [ B ] + z E [ N ] ) ÷ ( x + y + z ) where E[..] represent the total electronic energy including zero point correction. One can note that ΔE increases monotonically with the increase in n. This may suggest the increase in the stability of BN-structures with the increase in its dimension. As mentioned earlier, there is a charge transfer from B to H and from H to N atoms in BN-structures unlike CNNs in which there is always a charge transfer from terminal H to layered C atoms. It is interesting to analyze the effect of charge transfer and number of layers on the vibrational characteristics of BN-structures.

In Figs. 2, 3, we have plotted the vibrational IR spectra of BN-structures. NH stretching of BN-structures appears between 3580 and 3680 cm^–1whereas, BH stretching is obtained between 2625 and 2690 cm^–1. These IR active modes are pure, medium intense and approximately independent on the number of BN-layers. This is due to the fact that the increase in BN-layer does not affect the charge transfer significantly. This is reflected in the NBO charges on B and N in (4×1) and (4×2) BN-structures, which are approximately equal (see Fig. 4). The effect of the number of BN-layers appears in the region below 1400 cm^–1 which contains bending modes and torsions associated with BN-layer. In (4×1) single BN-layered structure, there are five distinguished peaks at 1367, 1260, 914, 686 and 437 cm^–1. The most intense peak at 1260 cm^–1 corresponds to the torsion in BN-layer. With the increase in BN-layer, lower frequency modes disappears which results in a single intense peak at approximately 1300 cm^–1 in (4×5) BN-structure.

It is also interesting to observe the frontier molecular orbital (FMO) energy gap of these BN-structures. The FMOs, such as the highest occupied MO (HOMO) and lowest unoccupied MO (LUMO), are the orbitals which participate in the chemical interaction. The FMO energy gap i.e. HOMO-LUMO gap can be regarded as a parameter to measure the chemical reactivity. The species with smaller HOMO-LUMO gaps are relatively more reactive and vice versa. The HOMO-LUMO gaps of BN-structures (Figs. 2, 3) are also listed in Table 1. These values follow the same trend as HOMO-LUMO gaps of corresponding CNNs [15]. However, the HOMO-LUMO gap of single layered BN-structure (6.75 eV) is much larger (almost twice) than that of corresponding CNN structure (3.5 eV) [15]. This can be expected due to the localization of lone pair of electrons on N atoms which causes distortion in (4×1) BN-structure from planarity.

4 Conclusions

In summary, our attempts to obtain BN analogs of CNNs do not meet a success. The resulting BN equilibrium structures are significantly distorted from hypothetical BNNN structures. In contrast to benzene like borazine and CNT like BNNT, it is not possible to realize CNN like BNNN. However, the distorted BN structures belong to at least local minima and their bond-lengths and HOMO-LUMO gaps follow the same trend as that of carbon nanoneedles. This distortion may be attributed to the repulsion created by lone pairs of N atoms and large electronegativity difference between B and N atoms which creates a polarity in BN bond. The study also demonstrates a case in which BN moiety fails to show a parallel behavior to CC bond.