Synthesis and optical property of large-scale centimetres-long silicon carbide nanowires by catalyst-free CVD route under superatmospheric pressure conditions

Abstract

Large-scale centimetres-long single-crystal β–SiC nanowires have been prepared using CH₄ as the carbon source and SiO or the mixture of Si and SiO₂ as the silicon source by a simple catalyst-free CVD route under superatmospheric pressure conditions. The nanowries grown on ceramic boat or corundum substrates, with lengths of several centimetres and the average diameters of around 40 nm, were composed of single-crystal β–SiC core along the [111] direction and amorphous SiO₂ shell of about 1–30 nm thick depending on the growth position along the flowing direction of the carrier gas. The total gas pressure is an important factor for the synthesis of the large-scale centimetres-long β–SiC nanowires, which can easily adjust the pressure of the vapors to supersaturation condition. The growth of the nanowires was governed by the Vapor-Solid mechanism. The β–SiC nanowires showed an intense blue light emission at room temperature.

---

Introduction

One-dimensional semiconductor nanomaterials, such as nanotubes, nanowires, nanobelts, nanorods and nanocables, have drawn a significant amount of attention because of their excellent electronic, physical and chemical properties.^1–4 The SiC nanowires (SiCNWs), as a wide band gap semiconductor, show wide applications in semiconductor, microelectronics and optoelectronics industry, nanoelectronics, field emission device and nanocomposites in many harsh conditions, such as that of high temperature, high power, and high frequency,^5–10 because of high breakdown electric field, high thermal conductivity, high resistance, high hardness, high strength, good field-emitting and unique optical properties.^11–15

Various synthetic methods, such as laser ablation,¹⁶carbon nanotubes-confined reaction,^17,18 high-frequency induction heating method,¹⁹chemical vapor deposition,^20,21 physical evaporation,²² and hydrothermal method,^23,24 have been used to grow SiCNWs. Most of these methods, however, involved complicated equipment and processes, vacuum conditions and metallic catalysts, which limit their further application. Especially, the SiCNWs synthesized by these methods have typically had lengths in the micrometre range. Only a few groups reported SiCNWs of several millimetres in length.^25–28 Ultra-long SiCNWs would be very important for the study of the mechanical, thermodynamic, and electric properties of nanostructures on a macroscopic scale, compared with their short counterparts.^29–31 Recently, Liet al. reported that large areas of centimetres-long or ultra-long SiCNWs have been prepared by pyrolysis of a polymer precursor with ferrocene or Fe(NO₃)₃ as catalyst by a CVD route.^32,33 However, the metallic catalyst remaining in the SiCNWs acts as an impurity and can cause degradation in the mechanical properties of nanocomposites, optoelectronics, etc. Although the ultra-long SiCNWs were also obtained by Liet al.³⁴ using polycarbosilane as the raw materials by a catalyst-free CVD route, the diameters fluctuate over the nanowires and the length in the millimetre range. Therefore, large area synthesis of centimetre-long SiCNWs without any catalyst is still a challenge.

According to the previous reports,^35,36carbon source has been regarded as an important factor, substantially affecting the rate of reaction and the morphology or size of synthesized SiCNWs. Up to now, carbon source such as active carbon,³⁷carbon nanoparticles,³⁸carbon nanotubes,^39,40CH₃SiCl₃ (MTS),^41,42 polysilacarbo-silane,^32–34C₃H₆^43,44 and CH₄,⁴⁵ and so on, was used to synthesize SiCNWs. At present, only a few groups reported the synthesis of SiCNWs using carbonaceous gas as the carbon source. For example, Liet al.⁴³ synthesized the SiCNWs by the CVD method using the C₃H₆ gas as the carbon source and the mixture of Si–SiO₂ as the silicon source at 1250 °C under a total gas pressure of 520–685 Torr. Liet al.⁴⁴ also synthesized the SiCNWs using the mixture of Si–SiO₂ and CH₄ as the raw materials at 1250–1200 °C following a decreasing atmospheric pressure from 600 to 200 Torr. Guo et al.⁴⁵ obtained the β–SiC nanowires with fins by chemical vapor deposition by using a powder mixture of milled Si and SiO₂ and gaseous CH₄ as the raw materials under a pressure of 200 Torr at 1350 °C. Yao et al.⁴⁶ reported the synthesis of 2H–SiC nanowhiskers by the reaction of silicon monoxide (SiO) with methane by the CVD route. During the growth, the temperature of furnace was kept at 1300 °C and the pressure was kept at 300 Torr. However, most of the SiCNWs were obtained below the pressure of 1 atm and the length in the micrometre range. Very few were synthesized under superatmospheric pressure. In addition, CVD is a suitable and widely used method to prepare nanowires. Thus far, there has been no report on the successful synthesis of centimetres-long SiCNWs in large areas by a catalyst-free CVD route under superatmospheric pressure.

In this paper, we will report the successful synthesis of centimetres-long single-crystal β–SiC nanowires in large areas by the starting materials of SiO or the mixture of Si and SiO₂ as the silicon source and CH₄ as the carbon source at 1350–1600 °C for 3 h by a simple catalyst-free chemical vapor deposition (CVD) route under superatmospheric pressure. Compared with the synthesis of SiCNWs reported previously, the total gas pressure during the process in our experiment is obviously higher than that of theirs, around 1.1–1.5 atm was used.

Results and discussion

After the CVD process, the boats (marked with (1) to (6) in Fig. 1 (a)) and the upper wall inside of the furnace (Fig. 1 (b)), even all over the tube (Fig. 1 (c)), were covered with a large amount of a light-blue or light-green or light-yellow or white or black cotton-like fibers, which depend on the location inside of the furnace, shown in Fig. 1 (a). The zone of fibers exceeds the flat-temperature zone of 27 cm to 38 cm, as shown by the stainless steel ruler in Fig. 1(a). Most of the fiber bundles obtained are up to several centimetres in length. Some of the light-green fibers were standing vertically in relation to the boats and the upper wall inside of the furnace, forming aligned arrays (Fig. 1 (d)). The height of the arrays was more than 1 cm and even to 3 cm. From the view of inside of the corundum tube (Fig. 1 (b) and (c)), we can prove that the length of the as-grown SiCNWs is up to 3 cm because the inner diameter of the corundum tube used in our experiment is 5 cm and the height of the ceramic boats is only 1.3 cm, as shown in Fig. 1(b) and (c). Meanwhile, from the side elevation and top view of boat (3) (Fig. 1(d) and (e)) and the fibers bundle reeled from the products (Fig. 1(f)), we also can prove that the length of the SiCNWs bundle is up to several centimetres.

Fig. 1 The digital photograph of, (a) the SiC nanowires grown on the ceramic boats of (1)–(7); (b) and (c) the inside of the tubular corundum furnace; (d) side elevation and (e) top view of the boat (3); (f) a bundle of fibers with a length up to several centimetres was reeled from the products.

To investigate the crystal structure and crystal type, the as-grown materials on a large scale were tested by XRD at room temperature. Fig. 2(a), (b) and (c) show the XRD patterns of the as-grown nanowires collected from the ceramic boats (1), (3) and (6), respectively, from which the diffracting peaks of the cubical β–SiC are detected. The major diffraction peaks in all patterns can be indexed as the (111), (200), (220), (311) and (222), respectively, reflecting a cubic β–SiC (unit cell parameter = 0.43589). These values are in accordance with the known values of β–SiC (JCPDS Card No. 29-1129). The strong and sharp peaks indicate that the SiCNWs had good crystallinity. From these patterns, we can also find that there is a strong broad peak centered at around 25.8 which can be indexed as the C (002) in Fig. 2(a), but it became very weak and broader in Fig. 2(b), and even disappeared in Fig. 2(c) and finally was replaced by a peak centered at around 21.6 which was attributed to the α–SiO₂. The phenomena are due to the higher concentration of the carbon vapor with respect to that of SiO vapor at the inlet while it is lower at the outlet. Therefore, at the inlet, the carbon vapor will not react completely and coat the surface of the SiCNWs displaying a black color, while, at the outlet, the SiO vapor will not react completely and coat the surface of the SiCNWs displaying a white or light-yellow color. At the centre, the carbon source vapor and the silicon source vapor almost react completely, so it displays a light-green color. This is the reason why the SiCNWs display a black or light-green or white color in a different ceramic boat. The above results show that large-scale centimetres-long single-crystal cubic β–SiC nanowires have been obtained by a simple catalyst-free CVD route, and indicate that the color of the SiCNWs depend on the position, varying from black, light-green, white to light-yellow color in turn.

Fig. 2 (a), (b) and (c) the XRD pattern of the centimetres-long SiC nanowires collected from ceramic boats of (1), (3) and (6), respectively.

The morphology of β–SiC nanowires is shown by SEM images. Fig. 3(a)–(b) show the typical low magnification and higher magnification field-emission scanning electron microscopy (FE-SEM) image of the black SiCNWs collected from the ceramic boat (1). Fig. 3(a) reveals that the product consists of wire-like structures and the density of the nanowires is very uniform. From the higher magnification image, we can see that the wires-like structure are composed of a big (around 80–200 nm) and rough shell and a small and smooth core (around 40 nm in diameter), as shown by red arrows in Fig. 3(b). The big and rough shell may be carbon nanoparticles. In order to improve the wrappages are carbon nanoparticles, a combustion experiment in air was done. After the treatment in air, the nanoparticles disappeared, as shown by the inset in Fig. 3(a). Fig. 3(c)–(d) show the typical low magnification and higher magnification FE-SEM image of SiCNWs collected from the ceramic boat (3). The low magnification scanning electron microscopy image (the inset in Fig. 3 (c)) indicated that some of the product consisted of aligned nanowires. The higher magnification reveals that the nanowires have a very smooth and straight morphology, which is different from the morphology of the product collected from the boat (1) completely. The average diameter is around 40 nm and has a length up to several centimetres. Fig. 3(e)–(f) show the typical low magnification and higher magnification FE-SEM image of SiCNWs collected from the ceramic boat (6). Some of the product collected from boat (6) also consisted of aligned nanowires, as shown in the inset in Fig. 3(e). The higher magnification reveals that the nanowires of boat (6) have a very silky and soft morphology with the average diameter around 60 nm, which is also different from the morphology of the product collected from the boat (1) and (3) completely.

Fig. 3 (a) Low-magnification and (b) higher-magnification FE-SEM image of SiC nanowires collected from ceramic boat of (1), the inset in (a) is the SEM image of the SiC nanowires after treatment in air; (c) low-magnification and (d) higher-magnification FE-SEM image of SiC nanowires collected from ceramic boat of (3); (e) low-magnification and (f) higher-magnification FE-SEM image of SiC nanowires collected from ceramic boat of (6).

To characterize the structure in detail, transmission electron microscopy (TEM) images were taken on a JEM-2000F transmission electron microscope. Fig. 4(a)–(b) show the typical TEM images of the as-grown SiCNWs collected from the ceramic boat (1), revealing that the periphery of the SiCNWs are very rough with a coating of carbon nanoparticles. Fig. 4(c)–(d) show the typical TEM images of the as-grown SiCNWs collected from the ceramic boat (3), revealing that the periphery of the SiCNWs are very smooth and clean with the average diameter around 40–50 nm, without any coating of nanoparticles, while Fig. 4(e)–(f) show the typical TEM images of the as-grown SiCNWs collected from the ceramic boat (6), revealing that the SiCNWs are composed of a small core and a thick shell forming the as-known core-shell structure. From Fig. 4(f), we can measure the thickness at around 10–30 nm and the core at around 5–15 nm. The chemical constitution of the SiCNWs collected from the boats (1), (3) and (6) are studied by EDX spectra. The EDX spectra are shown in Fig. 5(a), (b) and (c), respectively, revealing that all the SiCNWs are composed of silicon, carbon and oxygen, and no impurities exist. Further quantitative analysis results show that the oxygen in boat (6) is more than that in boat (1) and (3) while the carbon in boat (1) is more than that in boat (3) and (6). The results also indicate that the big and rough shell is composed of carbon nanoparticles.

Fig. 4 (a) Low-magnification and (b) higher-magnification TEM image of SiC nanowires collected from ceramic boat of (1); (c) low-magnification and (d) higher-magnification TEM image of SiC nanowires collected from ceramic boat of (3); (e) low-magnification and (f) higher-magnification TEM image of SiC nanowires collected from ceramic boat of (6).

Fig. 5 (a), (b) and (c) the EDX spectra of the centimetres-long SiC nanowires collected from ceramic boats of (1), (3) and (6), respectively.

In order to investigate the structure further, we performed a study using HRTEM images and selected area electron diffraction (SAED) measurements and select the SiCNWs from the ceramic boat (1), (3) and (6) as the representative material. Fig. 6(a) shows the HRTEM image of SiCNWs in boat (1). From it, a SiCNW with a high density of planar defects and stacking faults and a rough surface can be seen clearly. Fig. 6(b) is the amplified image of the zone covered by red dashed square in Fig. 6(a). The corresponding diffraction pattern obtained by fast Fourier transformation (FFT) is shown by the inset in Fig. 6(b). From the FFT image, we can calculate that the interplanar spacing is 0.252 nm, corresponding to {111} planes, indicating that the nanowires grow along the [111] direction. A high-resolution TEM image of boat (3) is shown in Fig. 6(c). We can see that the nanowires show a perfect lattice and a very thin shell (around 1 nm). A corresponding FFT image is shown by the inset in Fig. 6(c). After removing unwanted noise and enhancing periodic elements of an image through the technology of Fourier masking afforded in the Gatan, Inc., an inverse Fourier transformation was done. The corresponding image is shown in Fig. 6(d), from which we can see the image of atomic arrangement clearly. The lattice fringes can be seen clearly, with 0.252 nm of d spacing, corresponding to {111} planes, indicating that the nanowires also grow along the [111] direction. Fig. 6(e) shows the corresponding indexed selected-area electron diffraction (SAED) pattern of the nanowires in Fig. 4(d). The clear diffraction spots indicate that the nanowires were single crystals. The diffraction pattern was assigned to (111), (200), (220), (311) and (222), which are in accordance with the result of XRD patterns shown in Fig. 2. Fig. 6(f) shows the HRTEM image of a single SiCNW with a high density of planar defects and stacking faults. The inset is the corresponding SAED pattern. The SAED indicates a streaking pattern, which is a characteristic feature of a stacking fault structure. The HRTEM image of the SiCNWs in boat (6) is shown in Fig. 6(g). The FFT image (the inset in Fig. 6(g), corresponding to the zone covered by a square in Fig. 6(g)) and the corresponding inverse Fourier transformation image (in Fig. 6(h)) indicate that the nanowires were composed of a single-crystal β–SiC core (about 10–20 nm in diameter) along the [111] direction and amorphous SiO₂ shell (about 15–20 nm in thickness). This phenomenon may be due to the difference of the gas concentration along the flowing direction, which will be discussed in the following chapter.

Fig. 6 (a) HRTEM image of SiCNWs in boat (1); (b) the amplification image of the zone covered by red dashed square in (a), the inset is the corresponding FFT diffraction pattern; (c) HRTEM image of SiCNWs in boat (3), the inset is the corresponding FFT diffraction pattern of the specific region marked by the red dashed square in (c); and (d) the corresponding filtered inverse FFT image of the inset in (c); and (e) the SAED pattern; and (f) the high-magnification TEM image of single SiCNWs with planar defects and stacking faults, the inset is the corresponding SAED pattern; (g) HRTEM image of SiCNWs in boat (6), the inset is the corresponding FFT diffraction pattern of the zone covered by the red dashed square in (g); and (h) the corresponding filtered inverse FFT image of the inset in (g).

From the results of SEM, TEM and HRTEM, we can see that the morphology varies along the tube. The main reason is that the vapor concentration is different from the inlet to the outlet. Consequently, under superatmospheric pressure conditions, the flow rate of the vapors is very low along the direction of the tube, so that the carbon vapor concentration near the inlet is higher than that near the outlet. In addition, because the silicon sources are placed at the center of the tube and there is a thermal gradient between the center of tube and the mouth of pipe, the SiO vapor concentration near the outlet is higher than that near the inlet. Therefore, the C vapor near the inlet will be superfluous and absorbed on the surface of the SiCNWs, hence the black color of the SiCNWs, while the SiO vapor near the outlet will be superfluous and absorbed on the surface of the SiCNWs to form a thick amorphous SiO₂ shell layer, so that the SiCNWs near the outlet have cable-like morphologies and show a white color. At the center, the SiO and C vapor concentration are close to stoichiometric ratio, so the SiCNWs are composed of a single crystal SiC core and a thin amorphous SiO₂ shell and show a light-green color.

To further study the structure, we performed in situ Raman and FT-IR tests for the SiCNWs grown at the ceramic boats marked with (1), (3) and (6), respectively. Fig. 7(a), (b) and (c) show the Raman spectra of the as-synthesized β–SiC nanowires collected from the boats of (1), (3) and (6), respectively, at room temperature. Bulk β–SiC crystal usually has two characteristic peaks at around 796 and 972 cm⁻¹ corresponding to the first-order frequencies of modes of transverse optical (TO) and longitudinal optical (LO) phonons, respectively.^9,47 However, as can be seen from our measurement of the products, the TO mode rather than LO mode is detected. The sharp peaks at around 787.5, 787.8 and 786.6 cm⁻¹ shown in Fig. 7(a), (b) and (c) respectively correspond to the modes of transverse optical (TO) at the Γ point of the cubic β–SiC.⁴⁸ These peaks had red shifts of 8.5, 8.2 and 9.4 cm⁻¹ with respect to the TO mode of bulk SiC, respectively, which can be ascribed to the stacking faults in the nanowires.³² This is in accordance with the defects seen in the HRTEM images, as shown in Fig. 6(a), (b) and (f). The degradation of the LO mode is indicative of growth along the [111] direction by the select rule for zincblende structure polar crystal.^49,50 This result is in accordance with the results of XRD and HRTEM. In curve (a), two peaks centered at 1341.6 and 1587.2 cm⁻¹ were observed. Both of them correspond to the D and G bands of the graphitic phase, respectively, displaying a blue shift compared to the bulk graphite,⁵¹ indicating that a few carbon nanoparticles exist on the surface of the nanowires. Although, in curve (b), two weak peaks centered at 1522 and 1718 cm⁻¹ can also be seen clearly, they do not correspond to the D and G bands of the graphitic phase. Both of them may be attributed to the second-order frequencies of modes of transverse optical (TO) and longitudinal optical (LO) phonons. To our knowledge, up to now, no second-order modes of TO and LO of SiCNWs have been studied by Raman spectroscopy. This is the first time that the second-order TO and LO of SiCNWs with a diameter smaller than 100 nm has been studied by Raman spectroscopy. The peak centered at 1522 cm⁻¹ is the overtone of the TO (L) at 761 cm⁻¹, while the peak centered at 1718 is the overtone of the LO (M) at 859 cm⁻¹. Both of the peaks show a blue shift of 2–3 cm⁻¹ and 6 cm⁻¹ compared to the second-order TO and LO of the bulk 3C–SiC,^52–55 respectively. In curve (c), except the two peaks centered at around 1522 and 1718 cm⁻¹, another peak at around 926 cm⁻¹ ascribed to the axial optical mode of SiC polytypes was observed, as was done by Brioude's group and Li's group.^32,48 Different than the results obtained by Brioude's group and Li's group, this peak had red shifts of 14 cm⁻¹. With respect to the origin of the peak, there are different explanations for different groups. For example, Brioude's group⁴⁸ and Niu⁵⁶et al. attributed the peak centered at 925 to the LO modes, while Glinka⁵⁷et al. attributed the peak centered at 916 cm⁻¹ to the Raman peak of amorphous SiO₂. Based on the XRD, TEM, HRTEM and EDX analyses, we endorse the conclusion made by Glinka et al. and attribute the peak of 926 cm⁻¹ to the Raman peak of amorphous SiO₂. The Raman scattering spectra have confirmed that the as-grown centimetres-long SiCNWs are well crystalline.

Fig. 7 (a), (b) and (c) the Raman spectra of the centimetres-long SiC nanowires collected from ceramic boats of (1), (3) and (6), respectively.

The FT-IR transmittance spectra of the β–SiC nanowires are shown in Fig. 8. Curve (a), (b) and (c) are the FT-IR spectra of the products collected from the ceramic boats (1), (3) and (6), respectively. In curve (c), the strong peaks at 793.5 cm⁻¹ correspond to the stretching vibration of the Si–C bonds. The peak at 793.5 cm⁻¹ shows a 24.5 cm⁻¹ red shift compared with that of the bulk β–SiC, which may be ascribed to size confinement and surface effect.⁵⁸ The peaks at 477.8 and 1094.6 cm⁻¹are due to the stretching vibrations of Si–O,^59–61 which was ascribed to the amorphous silica shell located on the surface of the SiCNWs, as can be seen in the TEM images (Fig. 4(f)). In curve (b), only two peaks centered at 787.7 and 962.3, respectively, were shown, indicating that few amorphous silica shell and carbon were located on the surface, which is in accordance with the TEM image in Fig. 4(d). However, in curve (a), except for the peak of carbon and the OH group of the absorbed water, a very weak peak at around 796.5 cm⁻¹ of the stretching vibration of the Si–C bonds was also detected. No peaks ascribed to the stretching vibrations of Si–O can be seen. These indicate that the nanowires were coated with a thick shell of carbon which can weaken the signal of the Si–C bonds, as can be seen in the TEM images (Fig. 4(b)).

Fig. 8 (a), (b) and (c) the FT-IR spectra of the centimetres-long SiC nanowires collected from ceramic boats of (1), (3) and (6), respectively.

Fig. 9(a), (b) and (c) show the typical photoluminescence (PL) spectra of SiCNWs in the ceramic boats (1), (3) and (6), respectively. All the spectra were measured with an excitation wavelength of 380 nm at room temperature. As can be seen from Fig. 9(a), a UV-blue emission and a blue light emission centered at 405 nm and 429 nm, respectively, are observed. Both of the emission peaks are attributed to structure defects such as twins and stacking faults in the core of the nanowires and the special rough core-shell interface.^62–64 In Fig. 9(b), three strong and broad PL emission peaks were observed at the center wavelengths of 418, 442 nm and 458 nm, respectively. The violet light emission peak at 418 nm is ascribed to some intrinsic diamagnetic defect centers, such as the 2-fold-coordinated silicon long-pair centers (–O–Si–O–) and (–O–Si–C–O–).^65–68 Both the peak at 442 and 458 nm may be ascribed to the neutral oxygen vacancy (≡Si–Si≡) caused by the amorphous SiO₂ layer located on the surface of the SiCNWs, resulting from the forbidden transition from the triplet state to the singlet state at the neutral oxygen vacancy.^67–70 A main blue light emission peak centered at 463 nm and a shoulder peak at 442 nm are given by the SiCNWs in boat (6), as shown in Fig. 9(c). The blue light emission at 463 nm is attributed to neutral oxygen vacancy (≡Si–Si≡).⁷¹ Compared with the previously reported luminescence from bulk SiC, SiC nanospheres and films, the emission peaks of the SiCNWs are obviously blue-shifted,^72–76 which may be ascribed to quantum size effects of nanomaterials or the existence defects in amorphous layers and the special rough core-shell interface.^62–71 For the origin of these PL peaks, however, although the blue-shift is often thought to be caused by a size confinement effect, only such a SiC nanostructure with a size of 4 nm has an obvious quantum confinement effect, which is far smaller than 40–80 nm, the average diameter of the SiCNWs collected from the boats (1), (3) and (6), respectively. Therefore, the blue-shifts observed in this study must be caused by the existing oxygen defects in the amorphous SiO₂ shell, the special rough core-shell interface and structure defects such as twins and stacking faults in the core of SiCNWs.^63,64,72 Meanwhile, observing the PL spectra obtained from the boats (1), (3) and (6) carefully, we can find that the emission peak of 405 and 429 nm disappeared gradually and was replaced by some new peaks. Between the spectrum (b) and (c), red-shift was seen with increasing thickness of the SiO₂ shell layer. From the TEM and HRTEM images shown in Fig. 4 and Fig. 6, we can see that the surface morphology of the nanowires become smoother and smoother with the increase of the amorphous SiO₂ shell layer. Therefore, the imperfect parts in the core and at the core-shell interface caused by the defects are covered by the SiO₂ shell layer slowly, so that the peaks at 405 nm and 429 nm disappeared and the peaks at 418, 442 and 458 nm emerged gradually. With the further increase of the thickness of the SiO₂ shell, the intrinsic diamagnetic defects of the 2-fold-coordinated silicon long-pair centers (–O–Si–O–) and (–O–Si–C–O–) reduced while the neutral oxygen vacancy (≡Si–Si≡) increased, so that the peaks 418 and 442 nm disappeared and the peak 458 nm became the main emission and red-shifted to 463 nm. In addition, the SiCNWs in boat (3) were prepared at higher temperature than that in boat (1) and (6), thus better annealing is expected. The annealing can lower the density of defects, including twins and stacking faults. So the blue emission with red-shift from 458 to 463 nm remained, as the UV-blue emission at 405 nm and blue emission at 429 nm disappeared gradually. These may be the reasons why some emission peaks disappeared and some emission peaks red-shifted with increasing thickness of the amorphous SiO₂ shell layer. Of course, further research on the origin of the PL peak is necessary to clarify the origin of the PL shift.

Fig. 9 The PL spectrum of the SiNWs (a) in the ceramic boat (1); (b) in the ceramic boat (3); (c) in the ceramic boat (6).

The vapor–solid (VS),⁷⁷ vapor–liquid–solid (VLS),⁷⁸ solid–liquid–solid (SLS)⁷⁹ and oxide-assisted⁸⁰growth mechanisms have been developed for the formation of SiC nanostructures. In our work, no catalyst was used, and no catalyst particles were observed on the growth tip of the as-grown nanowires, which can be proven by the results of TEM, HRTEM and EDX. Therefore, the VLS and SLS growth mechanism can be ruled out since a landmark of intermediate nanoparticle is always seen at the tips of nanowires. On the basis of the experimental results, we can also use the vapor–solid growth mechanism to explain the growing process of SiCNWs obtained using SiO or the mixture of Si and SiO₂ as the silicon source, respectively. The detail is depicted as bellow: first, when the temperature reaches 1250 °C, a solid-solid reaction between the mixture of Si and SiO₂ will occur, which can be expressed in the following eqn (1):

Si_(solid) + SiO_2(solid) = 2SiO_(gas) (1)

If the SiO is used as the silicon source, the reaction (1) will not be necessary, and only a solid–vapor change will happen (SiO_(solid) → SiO_(gas)). Second, after reaching 1350–1600 °C in our experiment, CH₄ is introduced into the corundum tube, and the C vapor is obtained from decomposing CH₄, as indicated in eqn (2):

CH_4(gas) = C_(gas) + 2H_2(gas) (2)

When the pressure of SiO and C vapor rises to supersaturation conditions, reaction (3) and (4) as shown below would take place to decrease the system energy to form SiC nuclei on some active positions of the ceramic boats and the upper of the corundum tube by the vapor–solid mechanism.

SiO_(gas) + 2C_(gas) = SiC_(solid) + CO_(gas) (3)

SiO_(gas) + 3CO_(gas) = SiC_(solid) + CO_2(gas) (4)

The nuclei would preferentially grow in the [111] direction which has the lower energy than those of others in β–SiC. Because it has the highest atom density and the largest d spacing,⁶³ [111] oriented SiCNWs can be easily synthesized.

In fact, previously, some groups also synthesized SiCNWs using the same raw materials, but they could not obtain centimetre-long SiCNWs. What is the reason for that? On the basis of our experimental conditions, except the temperature conditions, we think the main effect factor is the total gas pressure. As mentioned above, the SiCNWs were synthesized under the low pressure conditions (below 1 atm) at relatively low temperatures of about 1200–1300 °C. However, the SiCNWs synthesized by us are obtained under superatmospheric pressure (1.1–1.5 atm) at 1350–1600 °C. According to the reported thermodynamic data,⁸¹ the calculated standard Gibbs free energy change of reaction (3) and (4) at 1400 °C is −200.4 kJ mol⁻¹ and 38.6 kJ mol⁻¹, respectively.⁸² We noted that at 1400 °C, the standard Gibbs free energy change of reaction (4) is positive, so that reaction (4) cannot proceed. Therefore, when the temperature is below 1400 °C, the SiC mainly formed according to eqn (3), which made the growth rate slow. So the yield and length will be limited. However, some researchers have confirmed that the reaction could occur under a supersaturated condition of CO vapor.^83,84

Previously, the SiCNWs were often synthesized by pumping the tube to a pressure below 1 atm. During the pumping process, a large amount of SiO, C, CO and CO₂ vapors in the tube were pumped out, so that the concentration of these vapors are low and the supersaturation condition cannot form easily under the lower pressure and temperature conditions. Consequently, the yield and length of SiCNWs are limited. However, in our work, the supersaturated condition was obtained by reducing the air displacement to adjust the total gas pressure to form the superatmospheric pressure. Based on the safety equipment requirement, we chose the total gas pressure between 1.1 and 1.5 atm. There are several advantages in inducing the air displacement. On one hand, it can prevent a large amount of SiO, C, CO and CO₂ vapors from running off and can heighten the concentration of the vapors and can raise the nucleation rate of the SiC nuclei. On the other hand, it benefits reaction (5) as shown below to react adequately, thus making the CO vapor reach the supersaturated condition quickly.

C_(gas) + CO_2(gas) = 2CO_(gas) (5)

In addition, as mentioned above, because of the superatmospheric pressure, the flow rate of the vapors along the tube is lower than that below 1 atm, so reactions (1)–(5) have enough time to proceed continuously. Therefore, the fast formation of the supersaturation conditions of SiO and CO make reactions (3) and (4) proceed easily even if the temperature is below 1400 °C. The higher unclear ratio and the sufficient reaction time would be beneficial for increasing the yield and making the SiCNWs grow longer and longer until finally reaching lengths of centimetres-long. To further verify the role of total pressure, an experiment was done at 1350 °C with the same other conditions. Centimeter-long SiCNWs were also obtained except with a slight decrease in the yield, which may be effected by the lower concentration of SiO vapor if the temperature is below 1400 °C. The result reveals that the total gas pressure is necessary for the formation of the centimetre-long SiCNWs by the catalyst-free CVD route. These may be the reasons why we can obtain the centimetre-long SiCNWs while other researchers cannot using the same raw materials.^43–46

Experimental section

Synthesis of centimetres-long SiCNWs

The SiO powder of about 6 g (purity: 99.999%, particle size: ∼60 μm, density: 2.1 g cm⁻³) or the mixture of Si powder (2 g) and SiO₂ (3 g) was divided into four identical sets and placed in four common ceramic boats (30 mm × 60 mm) (marked with (2), (3), (4), (5), as shown in Fig. 10(a)). Then, the boats were pushed into the center of a tubular corundum furnace (Φ = 55 mm, 1000 mm in length) with a flat-temperature zone of 27 cm, and one ceramic boat (marked with (1)) and two ceramic boats (marked with (6), (7)) without the raw materials were placed at the air inlet and air outlet, respectively, as shown in Fig. 10(a). After pumping down the furnace to a pressure of −0.1 MPa, a high purity argon gas was fed at a flow rate of 80 sccm into the furnace to maintain an inert atmosphere. When the pressure reaches 1.0 atm, the furnace was heated to the designed temperature according to the time–temperature curve, as shown in Fig. 10(b). Then a high purity CH₄ gas was fed into the furnace at a flow rate of 40 sccm after the temperature reaching the designed value and the pressure inside of the furnace was adjusted to 1.5 atm by the air outlet valve and maintained for reaction for 3 h. After finishing the reaction, the CH₄ was closed and the argon gas was fed in continuously and the furnace was cooled to 600 °C according to the time–temperature curve. During the process, the second thermal retardation was designed to ensure that the other style SiCNWs have enough time to change into β–SiC nanowires. Then the argon gas closed and the furnace was cooled to room temperature naturally. Finally, a large amount of light-blue or light-green or light-yellow or white or black cotton-like products with a length up to several centimetres were found covering the ceramic boats and the wall inside of the corundum tube (Fig. 1).

Fig. 10 (a) Schematic image of the preparation apparatus, and (b) the time–temperature curve.

Characterization

A JEOL JEM-2100F high-resolution transmission electron microscope (HRTEM) with 1.9 point-to-point resolution operating with a 200 KV accelerating voltage, equipped with energy dispersive X-ray spectroscopy (EDX INCA, OXFORD), and a FEI NanoSCI230 field emission scanning electron microscopy (FESEM) were used for the microstructure, morphological and chemical composition analyses. The FTIR spectra of the as-grown SiCNWs were obtained from a WQF-410 spectrometer in the transmission mode at room temperature with a resolution of 4 and a scan frequency of 32. Photoluminescence (PL) spectra were measured with an F-4500 fluorescence spectrophotometer with a Xe lamp at room temperature. The excitation wavelength was 380 nm. Raman spectra were measured with a Raman spectrometer (Renishaw 2000). An XRD was used for the crystallinity and crystal structure analyses.

Conclusions

In summary, we have successfully synthesized centimetres-long SiCNWs in a large amount by a simple catalyst-free CVD route under superatmospheric pressure conditions. The nanowires have a smooth surface and a straightforward body with a diameter in the range of 10–80 nm and the average diameter is around 40 nm. SEM, TEM, HRTEM, SAED, EDS, Raman, and FTIR analyses show that the nanowires are β–SiC nanowires and composed of a single crystal SiC core and an amorphous SiO₂ shell. The thickness of SiO₂ shell depends on the position inside of the furnace, along the flowing direction of Ar, the thickness increases gradually from around 1 nm to around 30 nm. The growth direction of the SiC nanowires is along the [111] direction. The total gas pressure is the major factor which would be beneficial for increasing the yield and the length. By the VS mechanism, we explained that the total gas pressure was responsible for the growth of large-scale centimetre-long SiCNWs under catalyst-free conditions. The SiCNWs showed a stable and intensive blue-light emission property with an excitation wavelength of 380 nm at room temperature.

Acknowledgements

This work was supported by Specialized Research Fund for the Doctoral Program of Higher Education (grant No. 20090162120008) and Natural Science Foundation of Hunan Province, China (grant No.09JJ3095).

References

1. C. M. Liber and Z. L. Wang, Mater. Res. Soc. Bull., 2007, 32, 99.
2. W. S. Shi, Y. F. Zheng, N. Wang, C. S. Lee and S. T. Lee, Adv. Mater., 2001, 13, 591.
3. Y. Cui and C. M. Lieber, Science, 2001, 291, 851.
4. J. Sha, J. J. Niu, X. Y. Ma, J. Xu, X. B. Zhang, Q. Yang and D. R. Yang, Adv. Mater., 2002, 14, 1219.
5. W. M. Zhou, L. J. Yan, Y. Wang and Y. F. Zhang, Appl. Phys. Lett., 2006, 89, 13105.
6. A. Fisher, B. Schroter and W. Richter, Appl. Phys. Lett., 1995, 66, 3182.
7. Q. G. Fu, H. J. Li, X. H. Shi, K. Z. Li, J. Wei and Z. B. Hu, Mater. Chem. Phys., 2006, 100, 108.
8. Z. W. Pan, H. L. Lai, C. K. A. Frederick, X. F. Duan, W. Y. Zhou, W. S. Shi, N. Wang, C. S. Lee, N. B. Wong, S. T. Lee and S. S. Xie, Adv. Mater., 2000, 12, 1186.
9. Z. C. Feng, A. J. Mascarenhas, W. J. Choyke and J. A. Powell, J. Appl. Phys., 1998, 64, 3176.
10. Y. Li, P. S. Dorozhkin, Y. Bando and D. Golberg, Adv. Mater., 2005, 17, 545.
11. E. W. Wong, P. E. Sheehan and C. M. Lieber, Science, 1997, 277(26), 1971.
12. X. T. Zhou, H. L. Lai, H. Y. Peng, C. K. F. Au, L. S. Liao, N. Wang, I. Bello, C. S. Lee and S. T. Lee, Chem. Phys. Lett., 2000, 318(1–3), 58.
13. W. Q. Han, S. S. Fan, Q. Li, W. J. Liang, B. L. Gu and D. Yu, Chem. Phys. Lett., 1997, 265(3–5), 374.
14. J. A. Powell, L. G. Matus and M. A. Kuczmarski, J. Electrochem. Soc., 1998, 134, 1558.
15. D. J. Larkin, J. Mater. Res. Soc. Bull., 1997, 22, 36.
16. W. S. Shi, Y. F. Zheng, H. Y. Peng, N. Wang, C. S. Lee and S. T. Lee, J. Am. Ceram. Soc., 2000, 83, 3228.
17. H. J. Dai, E. W. Wong, Y. Z. Lu, S. S. Fan and C. M. Lieber, Nature, 1995, 375, 769.
18. C. C. Tang, S. S. Fan, H. Y. Dang, J. H. Zhao, C. Zhang, P. Li and Q. Gu, J. Cryst. Growth, 2000, 210, 599.
19. F. L. Wang, L. Y. Zhang and Y. F. Zhang, Nanoscale Res. Lett., 2009, 4, 153.
20. J. C. Li, C. S. Lee and S. T. Lee, Chem. Phys. Lett., 2002, 355, 147.
21. H. J. Choi, H. K. Seong, J. C. Lee and Y. M. Sung, J. Cryst. Growth, 2004, 269, 472.
22. Z. S. Wu, S. Z. Deng, N. S. Xu, J. Chen, J. Zhou and J. Chen, Appl. Phys. Lett., 2002, 80, 3829.
23. Z. Ju, Z. Xing, C. Guo, L. Yang, L. Xu and Y. Qian, Eur. J. Inorg. Chem., 2008, 3883.
24. L. Z. Pei, Y. H. Tang, X. Q. Zhao and Y. W. Chen, J. Mater. Sci., 2007, 42, 5068.
25. K. F. Cai, A. X. Zhang and J. L. Yin, Nanotechnology, 2007, 18, 485601.
26. K. F. Cai, Q. Lei and A. X. Zhang, J. Nanosci. Nanotechnol., 2007, 7, 580.
27. K. F. Cai, Q. Lei and L. C. Zhang, J. Nanosci. Nanotechnol., 2005, 5, 1925.
28. K. F. Cai, A. X. Zhang, J. L. Yin, H. F. Wang and X. H. Yuan, Appl. Phys. A: Mater. Sci. Process., 2008, 91, 579.
29. C. Liu, H. Cheng, H. T. Cong, F. Li, G. Su, B. L. Zhou and M. S. Dresselhaus, Adv. Mater., 2000, 12, 1190.
30. Z. W. Pan, S. S. Xie, B. H. Chang, C. Y. Wang, L. Lu, W. Liu, W. Y. Zhou, W. Z. Li and L. X. Qian, Nature, 1998, 394, 632.
31. W. S. Shi, H. Y. Peng, Y. F. Zheng, N. Wang, N. G. Shang, Z. W. Pan, C. S. Lee and S. T. Lee, Adv. Mater., 2000, 12, 1343.
32. G. Y. Li, X. D. Li, Z. D. Chen, J. Wang, H. Wang and R. C. Che, J. Phys. Chem. C, 2009, 113, 17655.
33. G. Y. Li, X. D. Li, H. Wang and L. Liu, Appl. Phys. A: Mater. Sci. Process., 2010, 98, 293.
34. G. Y. Li, X. D. Li, H. Wang and L. Liu, Solid State Sci., 2009, 11, 2167.
35. S. Stolle, W. Gruner, W. Pitschke, L.-M. Berger and K. Wetzig, Int. J. Refract. Met. Hard Mater., 2000, 18, 61.
36. A. W. Weimer, K. J. Nilsen, G. A. Cochran and R. P. Roach, AIChE J., 1993, 39, 493.
37. S. Dhage, H. C. Lee, M. S. Hassan, M. S. Akhtar, C. Y. Kim, J. M. Sohn, K. J. Kim, H. S. Shin and O. B. Yang, Mater. Lett., 2009, 63, 174.
38. G. W. Meng, L. D. Zhang, C. M. Mo, S. Y. Zhang, Y. Qin, S. P. Feng and H. J. Li, Solid State Commun., 1998, 106, 215.
39. H. Dai, E. W. Wang, Y. Z. Lu, S. S. Fan and C. M. Lieber, Nature, 1995, 375, 769.
40. W. Q. Han, S. S. Fan, Q. Q. Li, B. L. Gu and D. P. Yu, Chem. Phys. Lett., 1997, 265, 374.
41. Q. G. Fu, H. J. Li, X. H. Shi, K. Z. Li, J. Wei and Z. B. Hu, Mater. Chem. Phys., 2006, 100, 108.
42. H. J. Choi, H. K. Seong, J. C. Lee and Y. M. Sung, J. Cryst. Growth, 2004, 269, 472.
43. Z. J. Li, H. J. Li, X. L. Chen, A. L. Meng, K. Z. Li, Y. P. Xu and L. dai, Appl. Phys. A: Mater. Sci. Process., 2003, 76, 637.
44. Z. J. Li, W. D. Gao, A. L. Meng, Z. D. Geng and L. Gao, J. Phys. Chem. C, 2009, 113, 91.
45. J. Z. Guo, Y. Zuo, Z. J. Li, W. D. Gao and J. L. Zhang, Physica E (Amsterdam), 2007, 39, 262.
46. Y. Yao, S. T. Lee and F. H. Li, Chem. Phys. Lett., 2003, 381, 628.
47. D. Olego and M. Cardona, Phys. Rev. B: Condens. Matter, 1982, 25, 3889.
48. M. Bechelany, A. Brioude, D. Cornu, G. Ferro and P. Miele, Adv. Funct. Mater., 2007, 17, 939.
49. Y. Ishikawa, A. V. Vasin, J. Salonen, S. Muto, V. S. Lysenko, A. N. Nazarov, N. Shibata and V.-P. Lehto, J. Appl. Phys., 2008, 104, 083522.
50. R. Kim, W. Qin, G. Wei, G. Wang, L. Wang, D. Zhng, K. Zheng and N. Liu, Chem. Phys. Lett., 2009, 475, 86.
51. R. P. Vidano, D. B. Fischbach, L. J. Willis and T. M. Loehr, Solid State Commun., 1981, 39(2), 341.
52. D. Olego and M. Cardona, Phys. Rev. B: Condens. Matter, 1982, 25, 1151.
53. W. Windl, K. Karch, P. Pavone, O. ütt, D. Strauch, W. H. Weber, K. C. Hass and L. Rimai, Phys. Rev. B: Condens. Matter, 1994, 49, 8764.
54. M. Hofmann, A. Zywietz, K. Karch and F. Bechstedt, Phys. Rev. B: Condens. Matter, 1994, 50, 13401.
55. J. C. Burton, L. Sun, F. H. Long, Z. C. Feng and I. T. Ferguson, Phys. Rev. B: Condens. Matter Mater. Phys., 1999, 59(11), 7282.
56. J. J. Niu and J. N. Wang, J. Phys. Chem. B, 2007, 111, 4368.
57. Y. D. Glinka and M. Jaroniec, J. Phys. Chem. B, 1997, 101, 8832.
58. G. W. Meng, Z. Cui, L. D. Zhang and F. Phillipp, J. Cryst. Growth, 2000, 209, 801.
59. V. Raman, O. P. Bahl and U. Dhawan, J. Mater. Sci., 1995, 30, 2686.
60. E. Bouillon, F. Langlais, R. Pailler, R. Naslain, F. Cruege, P. V. Huong, J. C. Sarthou, A. Delpuech, C. Laffon, P. Lagarde, M. Monthioux and A. Oberlin, J. Mater. Sci., 1991, 26, 1333.
61. X. Luo, W. Ma, Y. Zhou, D. Liu, B. Yang and Y. Dai, Nanoscale Res. Lett., 2010, 5, 252.
62. G. C. Xi, Y. Y. Peng, S. M. Wang, T. W. Li, W. C. Yu and Y. T. Qian, J. Phys. Chem. B, 2004, 108, 20102.
63. R. B. Wu, Y. Pan, G. Y. Yang, M. X. Gao, L. L. Wu, J. J. Chen, R. Zhai and J. Lin, J. Phys. Chem. C, 2007, 111, 6233.
64. R. B. Wu, B. S. Li, M. X. Gao, J. J. Chen, Q. M. Zhu and Y. Pan, Nanotechnology, 2008, 19, 335602.
65. J. Jang and H. Yoon, Adv. Mater., 2004, 16, 799.
66. Y. P. Guo, J. C. Zheng, A. T. S. Wee, C. H. A. Huan, K. Li, J. S. Pan, Z. C. Feng and S. J. Chua, Chem. Phys. Lett., 2001, 339, 319.
67. S. Kar and S. Chaudhuri, Solid State Commun., 2005, 133, 151.
68. H. Nishikawa, T. Shiryawa, R. Hakamura, Y. Ohike, K. Nagasawa and Y. Hama, Phys. Rev. B: Condens. Matter, 1992, 45, 586.
69. H. Nishikawa, T. Shiroyama, R. Nakamura, Y. Ohki, K. Nagasawa and Y. Hama, Phys. Rev. B: Condens. Matter, 1990, 41, 3794.
70. H. K. Seong, H. J. Choi, S. K. Lee, J. I. Lee and D. J. Choi, Appl. Phys. Lett., 2004, 85(7), 1256.
71. L. Liu, J. Y. P. Ko, M. J. Ward, Y. M. Yiu, T. K. Sham and Y. Zhang, J. Phys.: Conf. Ser., 2009, 190, 012134.
72. X. M. Liu and K. F. Yao, Nanotechnology, 2005, 16, 2932.
73. C. H. Liang, G. W. Meng, L. D. Zhang, Y. C. Wu and Z. Cui, Chem. Phys. Lett., 2000, 329, 323.
74. G. Z. Shen, D. Chen, K. B. Tang, Y. T. Qian and S. Y. Zhang, Chem. Phys. Lett., 2003, 375, 177.
75. J. Y. Fan, X. L. Wu and P. K. Chu, Prog. Mater. Sci., 2006, 51, 983.
76. K. A. Jachson, W. Schroter, Handbook of Semiconductor Technology, vol. 1, Wiley-VCH, Weinheim, 2000.
77. G. Yang, R. Wu, Y. Pan, J. Chen, R. Zhai, L. Wu and J. Lin, Physica E (Amsterdam), 2007, 39, 171.
78. D. F. Liu, S. S. Xie, L. J. C. Yan, F. Shen, J. X. Wang, Z. P. Zhou, H. J. Yuan, L. Gao, L. L. Song, F. Liu, W. Y. Zhou and G. Wang, Chem. Phys. Lett., 2003, 375, 269.
79. W. Y. Yang, Z. P. Xie, J. J. Li, H. Z. Miao, L. G. Zhang and L. N. An, Chem. Phys. Lett., 2004, 383, 441.
80. R. Q. Zhang, Y. Lifshitz and S. T. Lee, Adv. Mater., 2003, 15, 635.
81. Y. J. Liang, Y. C. Che, Handbook of Thermodynamics Data of Inorganic Substance , Dongbei University Press, Shenyang, 1993.
82. W. M. Zhou, Z. X. Yang, F. Zhu and Y. F. Zhang, Physica E (Amsterdam), 2006, 31, 9.
83. Y. H. Gao, Y. Bando, K. Kurashima and T. Sato, J. Mater. Sci., 2002, 37, 2023.
84. J. Wei, K. Z. Li, H. J. Li, Q. G. Fu and L. Zhang, Mater. Chem. Phys., 2006, 95, 140.

---