Systematic investigation of the ball milling–annealing growth and electrical properties of boron nitride nanotubes

Abstract

Boron nitride nanotubes (BNNTs) were grown on stainless-steel substrates by ball milling–annealing in an N₂/H₂ atmosphere. The effects of the precursor ratio and gas flow rate on the BNNT morphology were investigated. The correlations between experimental parameters and the BNNT morphology were interpreted based on tip- and base-growth modes. A study of the electrical properties of an individual BNNT showed that the contact resistances of the metal and BNNT contacts were improved by using an electroless technique. The main parameters such as resistivity, carrier concentration, and carrier mobility of the raw BNNTs suggest they have potential uses in high-power devices.

---

1 Introduction

Boron nitride nanotubes (BNNTs) are promising functional materials for use in a wide range of potential applications such as electronics, optoelectronics, and self-cleaning coatings. BN is a III nitride, therefore BNNTs have wide band gaps that are independent of the tube diameter, chirality, and number of tube walls.¹ They are therefore potential candidates for fabricating deep ultraviolet-light emission devices and optoelectronic devices.^2–6 The BNNT band gap can be tuned through doping, applying a transverse electric field, or covalent functionalization, and the electrical responses vary from insulating to semiconducting. Their typical semiconducting behavior broadens their use in nanometer-scale sensors, actuators, and advanced nanoelectromechanical systems. Huang et al.⁷ designed a submicrometer pH sensor based on biotin-fluorescein-functionalized multiwalled BNNTs with anchored Ag nanoparticles. Yu et al.⁸ fabricated a humidity sensor using a single Au-decorated BNNT as the sensing element. Lee reported a gigahertz oscillator based on double-walled BNNTs, which had higher frequencies than a carbon nanotube oscillator.⁹ Belonenko and Lebedev¹⁰ showed that BNNTs intercalated with alkali-metal atoms and alkaline-earth metal ions can be used as two qubit cells for quantum computers. They are also attractive for superhydrophobic applications because of their high surface : volume ratio and high curvature.^1,11 Other important applications of BNNTs include electrical insulators and gas storage because of the unique nature of their partial ionic bonds.¹² In addition, BNNTs have good chemical and thermal stabilities, therefore they can adapt to a large variety of rigorous chemical and thermal environments.¹²

All these potential applications are based on the unusual properties of BNNTs, which depend on the BNNT morphology.¹³ The synthesis of high-quality BNNTs with desired morphologies is therefore an important research area. Various methods have been used to synthesize BNNTs, e.g., arc discharge,¹⁴ laser ablation,^15,16 ball milling–annealing,^17–20 chemical vapor deposition (CVD),^21–24 self-propagation high-temperature synthesis,²⁵ and extended-pressure inductively coupled thermal plasma sintering.²⁶ Currently, CVD and ball milling–annealing are the methods most widely used for BNNT growth. Research has shown that the BNNT size and morphology can be modified by changing the growth parameters in a CVD system, e.g., the experimental setup, precursor type, precursor ratio, and temperature.²⁷ Pakdel et al. reported that the BNNT diameter increases with increasing temperature from 1200 to 1300 °C during CVD growth. Further increasing the temperature to 1400 °C can lead to secondary growth of tubes, which results in tube bundles with flower-like structures. In addition, low metal oxide : boron ratios result in the growth of BNNT films consisting of entangled curly tubes, whereas high metal oxide : boron ratios give thick tubes of diameters up to ∼350 nm with a semi-erect flower-like morphology.²⁸ Other studies have shown that changes in the growth duration and reaction atmosphere give BNNTs with different diameters and morphologies, and also lead to the selective formation of BN nanowires,²⁹ BN microtubes, and BN nanosheets.³⁰ The effects of the CVD growth parameters on the BNNT size and morphology have been systematically studied. However, there have been few studies of the effects of the growth parameters on the diameters and morphologies of BNNTs produced by milling–annealing. A precise understanding of the catalytic activity and gas flow rate would help to achieve controlled growth of BNNTs, which is a prerequisite for various potential applications. In this article, we report that one-dimensional BN nanostructures with different morphologies can be prepared selectively by controlling the catalyst concentration and N₂/H₂ flow rate in a milling–annealing system. The structures, morphologies, and elemental properties of BNNTs grown on stainless-steel substrates were studied in detail. The dependences of the growth mechanism and morphology on the catalyst concentration and N₂/H₂ flow rate were investigated using a vapor–liquid–solid (VLS) model.

A knowledge of the electrical properties of BNNTs is the first step in investigating their use in electronic/optoelectronic nanodevices, and considerable effort has been focused on experimental investigations of their electrical properties.^31,32 However, most studies were performed using a combination of transmission electron microscopy (TEM) and scanning tunneling microscopy. For the development of BNNT nanoelectronics, it is important to research their semiconducting properties and to achieve good contacts between BNNTs and electrodes.

BNNT-Based devices with low contact resistances were produced by an electroless technique using Ni on Al electrodes to improve the BNNT–metal contacts. This is a selective deposition technique, which does not require an additional mask or lithography. Two-terminal measurements were performed on the BNNTs produced using the proposed technique. The resistivity, carrier concentration, and carrier mobility of the raw BNNTs were deduced from the experimental current–voltage curves.

2 Experimental

Ball milling and annealing technique is used to synthesize boron nitride nanotubes with B and Fe(NO₃)₃·9H₂O as raw materials. All the precursors were of analytical grade (Sigma Aldrich) and used as received without further purification.

Before BNNT synthesis, the raw materials with different ratios were milled to produce fine precursor particles at meso- or nano-scale with enhanced chemical activity.³³ The ball milling process was performed in sealed milling vessels under 40 kPa nitrogen (N₂) gas with the weight ratios range from 30 : 1 : 0.08 to 30 : 1 : 4 of balls to B powder to Fe(NO₃)₃·9H₂O. And 1 g amorphous B powder was milled during each ball milling process. All the ball milling processes were performed for 20 h at 200 rpm of rotate speed under room temperature. 0.2 g of milled precursor powder was then dispersed into 0.4 ml deionized water to form B ink solution after 30 min ultrasonic treatment. The stainless steel substrates were painted with B ink solution, and followed by annealing process. The schematic of the annealing system used in this study is shown in Fig. 1(a). The stainless steel substrates were loaded into the center of the hot zone of the quartz tube, which is heated to the annealing temperature of 1200 °C under a N₂ flow of 200 sccm. Fig. 1(b) shows experimental procedures of gas change in the annealing process. During the annealing period, N₂/15% H₂ was carried into the hot zone, and the flow rate varied from 100 to 400 sccm. The annealing system was maintained at constant temperature and gas flow for 1 h. At the cooling period, N₂/15% H₂ flow was carried on until 550 °C. At room temperature, white color BNNTs were found deposited on stainless steel substrates.

Fig. 1 The schematic of the experimental set-up and experimental procedures of gas change in the annealing process: (a) experimental set-up; (b) the gas change with the temperature in the annealing process.

The as-synthesized samples were characterized with the aid of X-ray diffraction (XRD, X'Pert Pro MPD with Cu Kα radiation), Scanning Electron Microscope (SEM, Tescan VEGA 3 SBH), transmission electron microscopy (TEM, JEM 2100F at an accelerating voltage of 200 kV) and X-ray Photon Spectroscopy (XPS, Model: Thermo Scientific K-Alpha). The details are given in the following section. Electrical properties of an individual BNNT were measured by using an electrical parameter analyser HP4145 at room temperature. In order to form ohmic contact between metal electrodes and BNNT, electroless deposition of nickel onto aluminum electrodes was used to form electrically stable contacts.

3 Results and discussions

Product morphologies

The crystalline phase of as-synthesized samples was confirmed by XRD, as shown in Fig. 2. Several hexagonal BN peaks are visible in the XRD patterns of all the samples, besides, Fe–O compound peaks can be observed in the patterns, indicating that reactions involving the dissociation of Fe(NO₃)₃ took place during this experiment. This is consistent with the results of EDS spectrum analysis, as shown in Fig. 2(b). It is worth noting that when the mass ratio of Fe(NO₃)₃·9H₂O and B powders increases from 0.08 : 1 up to 4 : 1, the impurity peaks, such as Fe–O compound, Fe, and Fe–B compound, increase showing the residual catalyst in the reaction.

Fig. 2 XRD patterns and EDX spectrum of the samples: (a) XRD patterns of samples with 1 : 0.08 and 1 : 4 precursor's ratio; (b) EDX spectrum of samples grown with 1 : 0.08 precursor's ratio.

The elemental compositions of the as-synthesized BNNT samples were confirmed by XPS as well. After the binding energies were corrected for specimen charging by referencing the C 1s peak to 284.6 eV, the XPS spectra of the as-synthesized BNNTs with the different mass ratios of precursor's ratio are shown in Fig. 3. The survey shows the presence of B, N, C, O in all the samples. B 1s and N 1s peaks at 191.08 eV and at 398.08 eV, respectively. And the B/N atomic ratio is about 1.3, which corresponds to h-BN and is in good agreement with the available literature.^34,35 Compared with the samples grown with the small mass ratio of precursors 1 : 0.08, the Fe peak can be observed in the spectrum of 1 : 4 samples, indicating the 1 : 4 samples possess more catalyst impurities, which agrees with the results of XRD patterns.

Fig. 3 The XPS survey scan of the samples grown with 1 : 0.08 precursor's ratio (black curve) and 1 : 4 precursor's ratio (red curve).

In our experiments, BNNTs with different diameters and morphologies were obtained by using different ratios of precursors and different N₂/15% H₂ flow rates, as shown in Fig. 4. All the samples had non-uniform diameters. The results show that the mass ratios of the precursors and the gas flow rates both play a considerable part in nanotube synthesis, and affect the nanotube size and morphology. The mass ratio of the precursors also affects the nanotube yield. The average diameter and diameter distribution of the nanotubes both increased with increasing catalyst concentration. However, the length decreased greatly at catalyst concentrations higher than 1 : 0.5, and the nanotube quality deteriorated. A higher N₂/15% H₂ flow rate led to an increase in the average diameter of the grown nanotubes, except in the case of 100 sccm. When the gas flow was set at 100 sccm, the nanotube diameter range was broader than those of nanotubes prepared using flow rates of 200 and 300 sccm. When the gas flow rate was increased to 400 sccm, the diameter of the BNNTs on the uppermost BNNT film were large, in the range 200–600 nm. The nanotube morphology also changed, from curved to straight to curved, with increasing precursor mass ratio. The results of the present work differ from those reported for CVD systems,³⁶ presumably because of the differences between the precursors, temperatures, reaction atmospheres, and experimental set-ups.

Fig. 4 Effect of precursor's ratios and N₂/15% H₂ flow on the morphology of BNNT films, as revealed by SEM.

The products were further characterized using TEM. Fig. 5 shows the effects of the precursor ratio and gas flow rate on the BNNT morphology. The BNNTs had a bamboo-like morphology. The individual bamboo-like hollow-cored structures had a cup- or cone-type morphology, depending on the precursor ratio. At low catalyst concentrations, a cone-type bamboo-like hollow structure was obtained, and the average length of the hollow-cored structure increased with increasing precursor ratio. However, the cone angle of the bamboo-like structure decreased with increasing precursor ratio. The average wall thickness decreased from about 51.3 to 41.1 nm with increasing precursor ratio from 1 : 0.08 to 1 : 0.5. As the precursor mass ratio increased from 1 : 1 to 1 : 4, the BNNT bamboo-like structure became cup type, but irregular. The lengths of most of the hollow-cored structures decreased greatly and the bamboo-like structure changed to a bead-type morphology for the samples grown using a precursor mass ratio of 1 : 4. The average inner diameters of the bamboo-like structures of the BNNTs grown using precursor mass ratios of 1 : 1 to 1 : 4 were larger than those of the samples grown using precursor mass ratios of 1 : 0.08 to 1 : 0.5. Because of the larger average inner diameter, they have fewer walls (<100 walls, about 60–85 walls, as shown in the inset (a) of Fig. 5) compared with samples of similar diameter grown using precursor mass ratios of 1 : 0.08 to 1 : 0.5 (>100 walls, about 110–135 walls, as shown in the inset (b) of Fig. 5). It is important to note that for the samples grown using precursor mass ratios of 1 : 2 and 1 : 4, a few spindles containing catalyst nanoparticles inside the hemispherical ends were observed, as shown in the red circle region in Fig. 5.

Fig. 5 Effect of precursor's ratios and N₂/15% H₂ flow on the morphology of BNNT films, as revealed by TEM.

Unlike the precursor ratio, which affects the morphology and wall thickness, the gas flow rate affects the length of the hollow-cored structures. A higher flow rate of N₂/15% H₂ led to an increase in the average length of the hollow-cored structures, except in the case of 400 sccm. The samples grown at a gas flow rate of 400 sccm consisted of a few nanowires of large diameter.

Fig. 6 shows that quasi-spherical particles were predominantly observed at the tip-ends of the nanotubes. The components of these particles were identified based on TEM images and the corresponding energy-dispersive X-ray spectra of the tip-ends of the nanotubes (Fig. 6). The results show that BN layers encapsulate Fe particles at the centers of the tip-ends; B, N, Fe, Cr, Mg, and O are present at the tip-ends, suggesting catalytic growth of BNNTs.

Fig. 6 The TEM image and EDS spectrum of the quasi spherical particles at the tip-end of nanotubes.

Overview of the parameters' influence mechanism

Precursor ratio. It has been suggested that the growth of BNNTs produced using the ball milling and annealing technique is based on a VLS mechanism.^33,37,38 During annealing, the catalyst forms nanosized liquid or partially melted particles, which condense on the substrate when the partial vapor pressure is sufficient. The BN species diffuse into the catalyst particles attached to the substrate, leading to supersaturation and growth of BNNTs.^39–41 The nanotube diameter is believed to depend on the sizes of the catalyst particles and the partial pressure of the BN growth species.³⁹ Fig. 7(a) shows that a low catalyst concentration can result in smaller catalyst particles on the substrate, causing the formation of BNNTs with smaller diameters, as shown in Fig. 4. Although the catalyst concentration in the precursor was low, most of the nanotubes grew in a straight fashion, unlike the nanotubes grown by Pakdel.³⁶ Two possible reasons can be proposed to explain this. (1) A different substrate can result in a different nanotube morphology. (2) If the catalyst particles are small, the distribution density of the effective catalytic patches on the substrate is high, even at low catalyst concentrations. Due to high distribution density of the effective catalytic patches, BNNTs did not have enough space to grow, leading to growing in a straight fashion even under the presence of spatial fluctuations during growth species precipitation. As shown in Fig. 7(b), increasing the catalyst concentration increased the average size and distribution density of the catalyst particles because of catalyst nanoparticle aggregation, resulting in formation of straight BNNTs with larger diameters.⁴⁰ Fig. 7(c) shows that further increase in the catalyst concentration caused the catalytic nanoparticles to form into larger particles with small ones and peel off from agglomerates due to more nonuniform, leading to the formation of BNNTs with larger diameters and a broader diameter distribution. The formation of particles with larger diameters could also result in a lower distribution density of effective catalytic patches on the substrate, leading to BNNTs with a curly morphology, as shown in Fig. 4. In addition to spatial effects, structural defects could also result in curly BNNTs. X-ray diffraction (XRD) and X-ray photoelectron spectroscopy (XPS) showed that the product contained Fe when the precursor mass ratio was 1 : 4. Redundant catalyst results in more structural defects, leading to further growth of curly BNNTs.

Fig. 7 The proposed growth mechanism for BNNTs grown with different catalyst concentrations: (a) the lowest concentration; (b) medium concentration; (c) the highest concentration.

To further clarify the effect of catalyst concentration and to simulate our normal conditions just prior to BNNT growth, the catalyst particles were milled for 20 h, and then annealed at 1200 °C for 60 min without introducing boron. Scanning electron microscopy (SEM) was used to investigate the surface topography. The SEM images (Fig. 8) show that as the catalyst concentration increased from the lowest (Fig. 8(a)) to the highest (Fig. 8(c)), the most probable catalyst cluster size increased, and the size distribution broadened; this is in agreement with the results discussed above.

Fig. 8 The SEM images of catalyst with different concentration annealed at 1200 °C for 60 min: (a) the lowest concentration; (b) medium concentration; (c) the highest concentration.

A careful examination of the SEM images shows that all the BNNTs that we examined had catalyst particles at the tips. This can be clearly observed for the samples with a precursor mass ratio of 1 : 0.5, but is rare for a mass ratio of 1 : 0.08. The BNNTs grown using different catalyst concentrations had different morphologies. We attribute these differences to different catalytic growth modes. For the metal particles at the tips of the BNNTs grown with precursor mass ratios of 1 : 0.5 to 1 : 4, a tip-growth mode is suggested for the BNNTs prepared with high catalyst concentrations. Tubes grown using a low concentration (1 : 0.08) show a base-growth mode. At 1200 °C, some of the catalyst particles can anchor to the substrate through interdiffusion of metals.⁴² At a low catalyst concentration, almost all the metal particles are strongly attached to the substrate; small clusters are formed on the substrate, and remain anchored, during annealing. The base-growth mode may be involved in BNNT synthesis at the lowest catalyst concentration, therefore catalytic particles cannot be detected at the tip-ends. Increasing the catalyst concentration increases the average size of the catalyst clusters attached to the substrate. Metal particle are removed from the substrate during growth and are observed at the tip-ends of the BNNTs. The tip-growth model is the dominant growth mechanism for BNNTs.⁴³ When the catalyst concentration is further increased (1 : 4), the catalyst particles agglomerate and increase the distributions of both large and small diameters, based on the tip-growth model and base-growth model, separately.

Flow rate. Fig. 4 shows that the N₂/15% H₂ flow rate had an important influence on the morphology of the final product. An increase in the flow rate enhances the pressure in the reactor, resulting in an increase in the reaction rate, therefore more BN phases are generated and BNNTs with larger diameters are formed. As the flow rate increases to 400 sccm, BN fibers with large diameters can be observed on the uppermost BNNTs. A combination of the VLS and vapor–solid (VS) model can be used to explain the formation of large-diameter fibers. At a flow rate of 400 sccm, the local concentration of the reactant vapor would be too high for its complete consumption by VLS growth, leading to the formation of redundant vapor species, which could be deposited directly on the surfaces of the already formed fibers (VS process), resulting in production of thicker fibers.

The diameter of the BNNTs grown at 100 sccm is larger than that of the BNNTs grown at 200 sccm, and even larger than those of BNNTs grown at 300 sccm. This cannot be explained by the flow rate alone, but must be related to the N₂/15% H₂ ratio. As mentioned above, N₂ gas was initially introduced at room temperature to 900 °C to eliminate residual air in the quartz sintered tube, and it is assumed that N₂ remains in the sintered tube for a period of time after the flow has been switched off, similar to in the case of argon gas.^44,45 The residence time of N₂ gas in the sintered tube depends on the N₂/15% H₂ flow rate. At an N₂/15% H₂ flow rate of 100 sccm, the N₂ residence time would be longer than the other three flow rates, leading to an increase in the concentration of N₂, a higher actual flow rate, and a decrease in the H₂ content. The increase in the concentration of N₂ promotes the dissociation of N₂ gas to N radicals, which react with boron-containing species to form an h-BN phase.⁴⁶ In addition, it has been reported that the average diameter of BNNTs decreases with increasing H content,⁴⁷ therefore a high H₂ content could result in the formation of BNNTs with smaller diameters. We suggest that both factors result in the growth of BNNTs with larger diameters at an N₂/15% H₂ flow rate of 100 sccm.

Electrical property measurement

The investigation of electrical properties of an individual BNNT is vital and the first step towards to BNNT-based electronic/optoelectronic nanodevices. However, all the investigations should be based on good contacts between metal electrodes and nanotube. First, BNNTs were dispersed on silicon substrates with aluminum electrodes fabricated by a standard photolithography process. Then the contacts were formed by electroless deposition of Ni electrodes and post-processed by annealing at 500 °C. Current–voltage (I–V) characteristics were measured at room temperature with a maximum voltage of ±40 V before and after electroless Ni plating. A SEM image of the whole device is shown in the inset of Fig. 9(a). Additionally, the effect of annealing process on the contact resistance was investigated. Fig. 9(a) shows the I–V curves of an individual BNNT with diameter of 100 nm. The bias voltage was continuously scanned from −40 V to 40 V, with a step voltage of 1.0 V. The current is almost zero even at ±40 V high bias voltages before electroless plating, in comparison with that of after electroless plating and annealing. It can be explained that the surfaces of Al microelectrodes were oxidized to insulating Al₂O₃. In addition, the BNNTs deposited on the surface of microelectrodes only relying on the weak van der Waals force, resulting in a huge contact resistance. It is notable that without annealing, any current hardly transports the BNNT at ±40 V high bias voltages, even after electroless Ni plating. The adhesion of Ni layer should be taken into account. Before annealing, the adhesion of Ni layer is weak, so there may be a little gap between Ni layer and BNNT. The little gaps may result in the bad contact between Ni and BNNT, similar to the cold solder. Additionally, the BNNTs possess high resistance and wide band gap, which also impedes the transport of carriers through two huge contact barriers and the BNNT.

Fig. 9 I–V characteristic curves and ln I versus V curve of single BNNT: (a) I–V characteristic curves before and after electroless plating and annealing process; (b) ln I versus V curve after electroless Ni and annealing process.

After electroless Ni plating, the current shows a nonlinear increasing with the absolute value of the applied bias voltage due to the Schottky barriers between the BNNT and the electrodes. In addition, the I–V curves have perfect repeatability. When the voltage polarity of two electrodes reversed, the observed I–V characteristic curves are nearly symmetrical and the same as that before reversing voltage polarity. This is because current-induced local Joule heating can improve the contact properties, the perfect repeatability of I–V curves implies the existence of good and stable ohmic contacts between electrodes and BNNT. The resistance (R) and resistivity (ρ) of the individual BN nanotube obtained by differential computing the I–V characteristics curves are approximately equal to 9.5 × 10⁸ Ω and 312 Ω cm, respectively, which coincides with the resistivity in ref. 48. It is indicated that electroless technique can provide stable electrical and mechanical contacts between BNNTs and metallic electrodes.

In view of the fact that transport mechanism is dominated by tunnelling current under reverse bias in low-dimensional systems,⁴⁹ the reverse-biased Schottky barrier dominates the electrical transport under an intermediate bias.⁵⁰ And then, the total current (I) through a reverse-biased Schottky barrier dominating in the medium bias regime can be indicated by the following formula:^50–52

where J is the current density through a Schottky barrier, J_s is a slowly varying function of an applied bias, and S is the contact area associated with a barrier; in turn, E₀ is given by the equation: E₀ = E₀₀ coth(E₀₀/kT), where E₀₀ = ħq/2(p/m*ε)^1/2, p is hole concentration, m* is an effective hole mass of the as-synthesis BNNT, and ε is the dielectric constant. Because several studies have shown that as-synthesis undoped BNNTs are semiconducting, and p-doped, hole is considered in this letter.

With the aid of the I–V curves and formula (1), some parameters, such as hole concentration and carrier mobility can be obtained.^50–52 In the medium bias regime, the slope of q/(kT) − 1/E₀ in ln I versus bias voltage (V) is an approximately straight line, as shown in Fig. 9(b). Then the carrier concentration obtained via E₀ is approximately 5.79 × 10¹⁶ cm⁻³. The carrier mobility μ is estimated as ca. 0.35 cm² V⁻¹ s⁻¹ via the relationship of μ vs. ρ.

4 Conclusions

Multiwalled BNNT films were synthesized on stainless-steel substrates by milling–annealing using N₂/H₂ gas at 1200 °C. The effects of the catalyst concentration and N₂/H₂ flow rate on the growth, morphology, and structure of the BNNT films were studied systematically. The precursor ratio and gas flow rate both significantly affected the size and morphology of the nanotubes. The precursor ratio affected the nanotube yield. As the catalyst concentration increased, the mean diameter and diameter distribution of the nanotubes increased. However, when the catalyst concentration was higher than 1 : 0.5, the nanotube length decreased greatly, and the nanotube quality was poor because of the larger amount of residue. In addition, a N₂/15% H₂ flow rate led to an increase in the average diameter of the grown nanotubes, except in the case of 100 sccm, in which the N₂ content was higher. The electrical properties of a single BNNT sample were studied for the first time, using a two-terminal configuration. A transport mechanism based on a tunneling current, i.e., semiconductor behavior, was observed. Based on the current–voltage curves, the carrier concentration and mobility were estimated to be ca. 5.79 × 10¹⁶ cm³ and 0.35 cm² V⁻¹ s⁻¹, respectively, which are suitable for high-power devices. The formation of stable electrical contacts at the interfaces between the electrodes and BNNTs using the electroless Ni technique used in this study is an important step toward the use of BN one-dimensional nanostructures in electronic/optoelectronic nanodevices.

Acknowledgements

This work was financially supported by the National Natural Science Funds of China (No. 61404036), the National Basic Research Program of China (No. 2012CB934104), the Fundamental Research Funds for the Central Universities (No. HIT.NSRIF. 2015039, 01508536).

References

1. K. Yum and M. F. Yu, Measurement of Wetting Properties of Individual Boron Nitride Nanotubes with the Wilhelmy Method Using a Nanotube-Based Force Sensor, Nano Lett., 2006, 6, 329–333.
2. L. H. Li, Y. Chen, M. Y. Lin, A. M. Glushenkov, B. M. Cheng and J. Yu, Single deep ultraviolet light emission from boron nitride nanotube film, Appl. Phys. Lett., 2010, 97, 141104.
3. L. Li, L. H. Li, S. Ramakrishnan, X. J. Dai, K. Nicholas, Y. Chen, Z. Q. Chen and X. W. Liu, Controlling Wettability of Boron Nitride Nanotube Films and Improved Cell Proliferation, J. Phys. Chem. C, 2012, 116, 18334–18339.
4. C. H. Park, C. D. Spataru and S. G. Louie, Excitons and many-electron effects in the optical response of single-walled boron nitride nanotubes, Phys. Rev. Lett., 2006, 96, 126105.
5. H. Chen, Y. Chen, Y. Liu, C. N. Xu and J. S. Williams, Light emission and excitonic effect of boron nitride nanotubes observed by photoluminescent spectra, Opt. Mater., 2007, 29, 1295–1298.
6. J. Yu, D. H. Yu, Y. Chen, H. Chen, M. Y. Lin, B. M. Cheng, J. Li and W. H. Duan, Narrowed bandgaps and stronger excitonic effects from small boron nitride nanotubes, Chem. Phys. Lett., 2009, 476, 240–243.
7. Q. Huang, Y. Bando, L. Zhao, C. Y. Zhi and D. Golberg, pH sensor based on boron nitride nanotubes, Nanotechnology, 2009, 20, 415501.
8. Y. L. Yu, H. Chen, Y. Liu, L. H. Li and Y. Chen, Humidity sensing properties of single Au-decorated boron nitride nanotubes, Electrochem. Commun., 2013, 30, 29–33.
9. J. H. Lee, A study on a boron-nitride nanotube as a gigahertz oscillator, J. Korean Phys. Soc., 2006, 49, 172–176.
10. M. B. Belonenko and N. G. Lebedev, Two-qubit cells made of boron nitride nanotubes for a quantum computer, Tech. Phys., 2009, 54, 338–342.
11. L. H. Li and Y. Chen, Superhydrophobic Properties of Nonaligned Boron Nitride Nanotube Films, Langmuir, 2010, 26, 5135–5140.
12. J. S. Wang, C. H. Lee and Y. K. Yap, Recent advancements in boron nitride nanotubes, Nanoscale, 2010, 2, 2028–2034.
13. P. Ahmad, M. U. Khandaker, Y. M. Amin and Z. R. Khan, Influence of growth duration on size and morphology of boron nitride nanotubes grown via chemical vapor deposition technique, J. Phys. Chem. Solids, 2015, 85, 226–232.
14. N. G. Chopra, R. J. Luyken, K. Cherrey, V. H. Crespi, M. L. Cohen, S. G. Louie and A. Zettl, Boron nitride nanotubes, Science, 1995, 269, 966–967.
15. D. P. Yu, X. S. Sun, C. S. Lee, L. Bello, S. T. Lee, H. D. Gu, K. M. Leung, G. W. Zhou, Z. F. Dong and Z. Zhang, Synthesis of boron nitride nanotubes by means of excimer laser ablation at high temperature, Appl. Phys. Lett., 1998, 72, 1966–1968.
16. T. Laude, Y. Matsui, A. Marraud and B. Jouffrey, Long ropes of boron nitride nanotubes grown by a continuous laser heating, Appl. Phys. Lett., 2000, 76, 3239–3241.
17. Y. Chen, J. F. Gerald, J. S. Williams and S. Bulcock, Synthesis of boron nitride nanotubes at low temperatures using reactive ball milling, Chem. Phys. Lett., 1999, 299, 260–264.
18. L. Li, C. P. Li and Y. Chen, Synthesis of boron nitride nanotubes, bamboos and nanowires, Phys. E, 2008, 40, 2513–2516.
19. L. H. Li, Y. Chen and A. M. Glushenkov, Synthesis of boron nitride nanotubes by boron ink annealing, Nanotechnology, 2010, 21, 105601.
20. L. H. Li, Y. Chen and A. M. Glushenkov, Boron nitride nanotube films grown from boron ink painting, J. Mater. Chem., 2010, 20, 9679–9683.
21. S. Kalay, Z. Yilmaz and M. Culha, Synthesis of boron nitride nanotubes from unprocessed colemanite, Beilstein J. Nanotechnol., 2013, 4, 843–851.
22. R. Ma, Y. Bando and T. Sato, CVD synthesis of boron nitride nanotubes without metal catalysts, Chem. Phys. Lett., 2001, 337, 61–64.
23. R. Ma, Y. Bando, T. Sato and K. Kurashima, Growth, morphology, and structure of boron nitride nanotubes, Chem. Mater., 2001, 13, 2965–2971.
24. D. Golberg, Y. Bando, C. C. Tang and C. Y. Zhi, Boron nitride nanotubes, Adv. Mater., 2007, 19, 2413–2432.
25. J. L. Wang, L. P. Zhang, G. W. Zhao, Y. L. Gu, Z. H. Zhang, F. Zhang and W. M. Wang, Selective synthesis of boron nitride nanotubes by self-propagation high-temperature synthesis and annealing process, J. Solid State Chem., 2011, 184, 2478–2484.
26. A. Fathalizadeh, T. Pham, W. Mickelson and A. Zettl, Scaled synthesis of boron nitride nanotubes, nanoribbons, and nanococoons using direct feedstock injection into an extended pressure, inductively-coupled thermal plasma, Nano Lett., 2014, 14, 4881–4886.
27. P. Ahmad, M. U. Khandaker, Z. R. Khan and Y. M. Amin, Synthesis of boron nitride nanotubes via chemical vapour deposition: a comprehensive review, RSC Adv., 2015, 5, 35116–35137.
28. A. Pakdel, C. Y. Zhi, Y. Bando, T. Nakayama and D. Golberg, A comprehensive analysis of the CVD growth of boron nitride nanotubes, Nanotechnology, 2012, 23, 215601.
29. P. Ahmad, M. U. Khandaker, Z. R. Khan and Y. M. Amin, A simple technique to synthesize pure and highly crystalline boron nitride nanowires, Ceram. Int., 2014, 40, 14727–14732.
30. P. Ahmad, M. U. Khandaker, Y. M. Amin and Z. R. Khan, Synthesis of boron nitride microtubes and formation of boron nitride nanosheets, Mater. Manuf. Processes, 2014, 30, 184–188.
31. R. Arenal, M. S. Wang, Z. Xu, A. Loiseau and D. Golberg, Young modulus, mechanical and electrical properties of isolated individual and bundled single-walled boron nitride nanotubes, Nanotechnology, 2011, 22, 265704.
32. H. M. Ghassemi, C. H. Lee, Y. K. Yap and R. S. Yassar, Field emission and strain engineering of electronic properties in boron nitride nanotubes, Nanotechnology, 2012, 23, 105702.
33. L. Li, S. D. Fan, C. C. Zhuang, C. C. Ban and X. W. Liu, Customizable local superhydrophobic micro channels by BNNTs growth, Microelectron. Eng., 2015, 141, 223–227.
34. G. Wen, T. Zhang, X. X. Huang, B. Zhong, X. D. Zhang and H. M. Yu, Synthesis of bulk quantity BN nanotubes with uniform morphology, Scr. Mater., 2010, 62, 25–28.
35. J. L. Wang, Z. L. Li, Y. L. Gu, X. W. Du, Z. X. Zhang, W. M. Wang and Z. Y. Fu, Synthesis of boron nitride nanotubes using glass fabrics as catalyst growth framework, Ceram. Int., 2015, 41, 1891–1896.
36. A. Pakdel, C. Y. Zhi, Y. Bando, T. Nakayama and D. Golberg, A comprehensive analysis of the CVD growth of boron nitride nanotubes, Nanotechnology, 2012, 23, 215601.
37. L. Li, L. H. Li, Y. Chen, X. J. Dai, T. Xing, M. Petravic and X. Liu, Mechanically activated catalyst mixing for high-yield boron nitride nanotube growth, Nanoscale Res. Lett., 2012, 7, 1–8.
38. L. Li, X. Liu, L. Li and Y. Chen, High yield BNNTs synthesis by promotion effect of milling-assisted precursor, Microelectron. Eng., 2013, 110, 256–259.
39. C. H. Lee, J. S. Wang, V. K. Kayatsha, J. Y. Huang and Y. K. Yap, Effective growth of boron nitride nanotubes by thermal chemical vapor deposition, Nanotechnology, 2008, 19, 455605.
40. A. Pakdel, C. Y. Zhi, Y. Bando, T. Nakayama and D. Golberg, A comprehensive analysis of the CVD growth of boron nitride nanotubes, Nanotechnology, 2012, 23, 215601.
41. A. T. Matveev, K. L. Firestein, A. E. Steinman, A. M. Kovalskii, O. L. Lebedev, D. V. Shtansky and D. Golberg, Boron nitride nanotube growth via boron oxide assisted chemical vapor transport-deposition process using LiNO₃ as a promoter, Nano Res., 2015, 8, 2063–2072.
42. V. L. Merkulov, A. V. Melechko, M. A. Guillorn, D. H. Lowndes and M. L. Simpson, Alignment mechanism of carbon nanofibers produced by plasma-enhanced chemical-vapor deposition, Appl. Phys. Lett., 2001, 79, 2970–2972.
43. A. Gohier, C. P. Ewels, T. M. Minea and M. A. Djouadi, Carbon nanotube growth mechanism switches from tip-to base-growth with decreasing catalyst particle size, Carbon, 2008, 46, 1331–1338.
44. D. Özmen, N. A. Sezgi and S. Balcı, Synthesis of boron nitride nanotubes from ammonia and a powder mixture of boron and iron oxide, Chem. Eng. J., 2013, 219, 28–36.
45. P. Ahmad, M. U. Khandaker, Y. M. Amin, N. Muhammad, A. R. Uaman and M. Amin, The effect of reaction atmosphere and growth duration on the size and morphology of boron nitride nanotubes, New J. Chem., 2015, 39, 7912–7915.
46. J. Yu, Y. Chen, R. Wuhrer, Z. W. Liu and S. P. Ringer, In Situ Formation of BN Nanotubes during Nitriding Reactions, Chem. Mater., 2005, 17, 5172–5176.
47. J. Yu, B. C. P. Li, J. Zhou and Y. Chen, Influence of nitriding gases on the growth of boron nitride nanotubes, J. Mater. Sci., 2007, 42, 4025–4030.
48. C. C. Tang, Y. Bando, Y. Huang, S. Yue, C. Gu, F. Xu and D. Golberg, Fluorination and Electrical Conductivity of BN Nanotubes, J. Am. Chem. Soc., 2005, 127, 6552–6553.
49. J. Appenzeller, M. Radosavljević, J. Knoch and P. Avouris, Tunneling Verses Thermionic Emission in One-Dimensional Semiconductors, Phys. Rev. Lett., 2004, 92, 048301.
50. X. d. Bai, D. Golberg, Y. Bando and C. y. Zhi, et al., Deformation-Driven Electrical Transport of Individual Boron Nitride Nanotubes, Nano Lett., 2007, 7, 632–637.
51. F. A. Padovani and R. Stratton, Field and thermionic-field emission in Schottky barriers, Solid-State Electron., 1966, 9, 695–707.
52. Z. Y. Zhang, C. H. Jin, X. L. Liang, Q. Chen and L. M. Peng, Current–voltage characteristics and parameter retrieval of semiconducting nanowires, Appl. Phys. Lett., 2006, 88, 073102.

---