1 Introduction

SnO2 is a promising key functional material for a wide range of practical applications. It is widely used as a gas sensor, an electrode material for energy conversion and storage applications, and has an application in field emission and photoluminescence [1–4]. SnO2 nanocrystals, hollow spheres, nanowires, nanobelts, and nanotubes have been synthesized by hydrothermal, inverse microemulsion, thermal evaporation, and the redox reaction mechanism [5–8]. There is newly emerging interest in one-dimensional nanostructured SnO2 materials such as nanobelts, nanotubes, and nanowires, in addition to the extensively studied SnO2 films and nanoparticles. Recently, attempts have been made to fabricate SnO2 nanobelts and nanowires for use in sensors for detecting various gases such as carbon monoxide, nitrogen dioxide, and ethanol vapor [9–11]. SnO2 nanowires have also been researched for application in lithium-ion batteries [12].

Core–shell-structured materials have attracted much attention in physics, chemistry, and material science communities because of their specific structures and unique properties [13, 14]. Advanced materials derived from core–shell composite particles are of extensive scientific and technological interest because of fine-tuning material properties [15]. Wang and his collaborators synthesized a composite of SnO2 nanoparticles coated on SiO2 microspheres and found a marked blueshift phenomenon [16]. One-dimensional (1D) heterostructured nanomaterials have attracted great attention due to their novel and distinct properties, and more effort has been taken to explore possible high-performance nanodevices based on them [17, 18]. Synthesis of SnO2–CuO heterojunction using electrospinning has been reported for application in detecting CO [19]. All inorganic frameworks of tin dioxide shell as cathode material improve cycle performance for lithium–sulfur batteries [20]. Xu et al. improved the mechanical integrity and three-dimensional electron transport for lithium batteries by heterogeneous branched core–shell SnO2–PANI nanorods [21].

So far, a 1D SnO2–SiO x core–shell heterostructured nanorod has not been reported in the literature. SnO2 and SiO x show high storage capacities as electrode materials [2, 22], but the high storage capacities fade quickly with cycling due to severe volume expansion. Heterostructured materials have been shown in other systems to prevent the electrochemical degradation induced by enormous volume changes during Li+ insertion or extraction [23]. In this paper, we report a one-step preparation method for SnO2 nanorods with a coaxially grown SiO x overlayer and the vertical selective growth is accomplished by adjusting the kinetic parameters. The morphology and growth process of the 1D free-standing core–shell heterostructure are also discussed in detail.

2 Experimental methods

The synthesis system used here includes a high-temperature horizontal tube furnace. A quartz tube with an outer diameter of 50 mm and a length of 1200 mm was placed inside the furnace. Iodine powder 2–6 g in an alumina boat was placed upstream of the growth site in the quartz tube, just outside of the tube furnace. Sn particles were used as the evaporation source. The evaporation materials were placed in an alumina boat and loaded into the center of the tube furnace. At high temperatures T = 1050 °C, sublimed iodine reacts with Sn to produce a gaseous product of metal iodide that is carried downstream by argon gas flow to a different temperature zone 900–1000 °C, where it reacts with oxidized Si substrates. Argon (99.9 % pure) was made to flow at atmospheric pressure through the reaction tube at a high rate before synthesis and then reduced to a lower flow rate of 10–60 SCCM during synthesis (SCCM denotes cubic centimeter per minute at standard temperature and pressure).

Oxidized Si substrates were produced from silicon wafer chips roughly etched using a wet oxidation treatment: Si substrates were heated in a stirred “metal etch” solution (30 % H2O2:37 % HCl:H2O; 1:1:5 v/v) at 70 °C for 20 min, followed by rinsing in deionized water and air-drying.

The morphology of the as-grown products was examined using a field emission scanning electron microscope (SEM, HITACH S-4800). Transmission electron microscopy (TEM) samples were prepared by scripting as-grown substrate using copper grids. The selected area electron diffraction (SAED) patterns and high-resolution transmission electron microscope (HRTEM) images were obtained using a JEOL JEM-2010 microscope.

3 Results and discussion

A typical process for synthesizing coaxial SnO2–SiO x heterostructured nanorods is described as follows. The center temperature of furnace is fixed at 1050 °C, and oxidized Si substrates are placed downstream where the temperature is about 950 °C. Flow of argon was controlled at a rate of 30 SCCM. The free-standing nanorods and nanotubes shown in Fig. 1 were prepared via the above process. As shown in Fig. 1a, a large number of nanorods randomly grow on the oxidized substrates. As-grown depositions show a uniform morphology characterized by multi-prism shape, smooth surface, a diameter of 100–200 nm, and a length of 1–2 μm, as presented in Fig. 1a, b. The multi-prism morphology is similar to the SnO2 nanorod morphology of another report [24]. The nanorods grow from the substrates and are not simply deposits on the substrates.

Fig. 1
figure 1

Morphology of nanorods under typical synthesis process

Full size image

The element line distribution of a single nanorod is shown in Fig. 2. There are composition fluctuations of three elements (Si, Sn, and O) as shown in Fig. 2. Though Fig. 2 shows a reduction in Si content, this apparent reduction is related to a strong Si background from the substrate. It can be concluded from these data that the compositions of the nanorods may include Si, Sn, and O.

Fig. 2
figure 2

Element line distribution of nanorods under typical synthesis process

Full size image

Transmission electron microscopy further confirms that the nanorod appears to have a smooth surface, but is covered with a shell layer. The diameter of the nanorod measured by TEM is similar to the result of above SEM observation. From the selective diffraction pattern in inset of Fig. 3a, it can be seen that a characteristically amorphous halo pattern and a crystalline diffraction pattern occur simultaneously. The crystalline diffraction pattern is difficult to index due to screening from the amorphous layer. To verify core–shell substance phase and composition, selective diffraction patterns and energy spectra should be characterized after removing the amorphous layer. Figure 3b shows the TEM photograph of a nanorod with part of the shell layer removed. An amorphous SiO x shell layer (area labeled as 1 in Fig. 3b) is verified by energy spectrum shown in Fig. 3c. As shown in Fig. 3d, there is no Si element in energy spectrum of core layer (zone labeled as 2 in Fig. 3b). From these data, it may be concluded that the phase of core layer material should be tin oxide. The chemical composition from energy spectrum corresponds with the element line distribution results from SEM.

Fig. 3
figure 3

TEM image and energy spectrum of a nanorod

Full size image

In order to analyze phase and microstructure of the core layer, we observe the lattice fringe phase and selective diffraction pattern for 2 zones in Fig. 3b. The core material phase is concluded as SnO2 by the selective diffraction pattern shown in inset of Fig. 4. Spacing between (101) planes in SnO2 crystal is corresponding to spacing (2.65 nm) of lattice fringe phase shown in Fig. 4, so the growth direction of nanorod is the [101].

Fig. 4
figure 4

Lattice fringe phase, selected diffraction of SnO2 core

Full size image

According to the theoretical calculation of SnO2 crystal surface energy, the sequence of surface energy per crystal face is (110) < (100) < (101) ≪ (001) [25]. The (110) facet has the lowest surface energy, and [110] should be a reasonable growth direction according to the thermodynamical stability. Theoretical calculation of SnO2 crystal surface energy originates from the perfect crystal, but there are different defects in different processes. Those defects could affect the value of crystal surface energy. Theoretical calculation and experimental evidence all suggest that the reduced (101) surface has a lower surface energy than the (110) surface [26, 27]. In our process, low oxygen partial pressure and high reaction temperature are advantageous to form the reduced (101) surface. So the [101] growth direction in our process also satisfies the lowest energy principle in thermodynamics.

The process for synthesizing free-standing SnO2–SiO x core–shell nanorods includes chemical vapor transport, nucleation, and growth. Probable reactions are shown as follows.

$${\text{Sn}}({\text{s}}) \to {\text{Sn}}({\text{g}})$$
(1)
$${\text{Sn}}({\text{g}}) + {\text{I}}({\text{g}}) \to {\text{SnI}}_{2} ({\text{g}})$$
(2)
$${\text{SiO}}_{x} ({\text{s}}) \to {\text{Si}}({\text{s}}) + {\text{SiO}}_{2} ({\text{s}})$$
(3)
$${\text{SiO}}_{2} ({\text{s}}) + {\text{SnI}}_{2} ({\text{g}}) \to {\text{SiO}}_{x} ({\text{s}}) + {\text{SnO}}({\text{s}}) + {\text{I}}_{ 2} ({\text{g}})$$
(4)
$${\text{SnO}}({\text{s}}) + {\text{SiO}}_{x} ({\text{s}}) \to {\text{SiO}}_{y} ({\text{s}}) + {\text{SnO}}_{2} ({\text{s}})$$
(5)

Tin easily evaporates at 1050 °C (as shown in reaction 1), and it can react with iodine vapor upstream (as shown in reaction 2). Then gaseous product (SnI2) was transported to the growth zone for nucleation and growth under Ar flow. The amorphous SiO x layer appears on the substrate after wet-etching, and then dismutation reaction (reaction 3) occurs for SiO x at high temperature [28]. The SiO2 microzone produced by reaction 3 provides a site of heterogeneous nucleation. At the nucleation site, the reactions about SnI2 and SiO2 may occur according to reactions 4 or 5. A core–shell (SnO2–SiO x ) structure should thus be synthesized due to the phase separation in order to reduce interface energy between the two phases [29]. The extremely precise control of core–shell size and thickness requires a better understanding of the core–shell growth mechanism, and further experiment is underway in our research group.

In order to clarify the kinetic process of synthesizing free-standing nanorods, the synthesis conductions (temperature, gas flow, state of substrate, and I2 content parameters) are tuned. The reaction and deposition temperature directly affects the chemical reaction and surface adsorption and thus changes the morphology of deposited materials. Surface adsorption should be performed first before deposition; the adsorption ability can be described as follows.

$$\sigma = \tau F$$
(6)
$$\tau = \tau_{0} \exp [\Delta H_{\text{ads}} /{\text{RT}}]$$
(7)
$$F = 3.5 \times 10^{22} \frac{P}{{\sqrt {\text{MT}} }}$$
(8)

where \(\sigma\), \(\tau\), \(F\), \(\tau_{0}\), \(\Delta H_{\text{ads}}\), \(R\), \(T\), \(P\), \(M\) are the surface coverage, the average residence time, the gas incidence flux, constant relation with vibration period, adsorption heat, gas constant, absolute temperature, vapor pressure of adsorbed molecular, and relative molecular mass.

In the case of other constant parameters, higher (1150 °C) and lower (950 °C) temperatures result in no obvious product deposition. Though higher temperature is beneficial for promoting chemical reactions, it is also an obstacle for surface adsorption. The limited surface adsorption and the limited reaction cases occur at high temperature and low temperature, respectively. Deposition therefore only occurs at the proper temperature.

At proper temperature, the morphology of deposition is also directly controlled by the mass transfer coefficient. In the open chemical vapor deposition system, gas flow near substrates may be considered as fitting the laminar flow model. Diffusion is always the dominant transport mechanism near the surface of a solid substrate. According to the laminar model, the mass transfer coefficient may be described as follows [30].

$$h_{\text{mi}} = 0.664\frac{D}{l}\text{Re}_{x}^{1/2} {\text{Sc}}^{1/3}$$
(9)

where \(h_{\text{mi}}\), \(l\), \(\text{Re}_{x}\), \({\text{Sc}}\), \(D\) are the mass transfer coefficient, characterized length, Reynolds number, Schmidt number, and diffuse coefficient.

After deducing,

$$h_{\text{mi}} = 0.664\left(\frac{v}{l}\right)^{1/2} \left(\frac{\rho }{\eta }\right)^{1/6} D^{2/3}$$
(10)

where \(h_{\text{mi}}\), \(v\), \(l\), \(\rho\), \(\eta\), \(D\) are the mass transfer coefficient, the velocity of gas flow, characterized length, fluid density, viscosity, and diffuse coefficient.

The mass transfer coefficient may be controlled by adjusting gas flow. When flow was adjusted to a lower value (10 SCCM), there is no obvious disposition on substrates due to an insufficiently large mass transfer coefficient. When gas flow was controlled at 30 SCCM, free-standing nanorods were deposited on substrate (shown in Fig. 1). Adjusting gas flow to the bigger value (60 SCCM), as-grown depositions on substrate appear with shrimp-like, flower-like, and worm-like morphologies (shown in Fig. 5a, b). The vertical growth and horizontal growth models are now presented according to the morphology difference shown in Figs. 1 and 5. The limited resource was transferred to substrate under a proper gas flow (30 SCCM). The limited resource was selectively adsorbed on the growth nucleus, so free-standing nanorods appear on substrate. This selective grow process is considered as the vertical growth model. Under a greater gas flow, more resource was transferred to substrate due to the larger mass transfer coefficient. The surplus resource accelerates the horizontal growth of nanomaterial. The horizontal growth model is similar to the initial stage of film deposition [31].

Fig. 5
figure 5

Deposited materials morphology under gas flow of 60 SCCM

Full size image

State of substrate and content of I2 also affect nucleation and chemical transport during synthesis, respectively. There is no obvious substance deposition without I2 or with untreated Si substrate.

4 Conclusions

The free-standing coaxial SnO2–SiO x core–shell structure nanorods were synthesized by the CVT method. The coaxial SnO2–SiO x heterostructured nanorods appear with uniform morphology, smooth surface, a diameter of 100–200 nm, and a length of 1–2 um. The low oxygen partial pressure and high reaction temperature in our process make the reduced (101) surface, which has a lower surface energy than the (110) surface. So the [101] growth direction is also reasonable according to the lowest energy principle in thermodynamics. The SiO x dismutation reaction provided the nucleation site, and formation of core–shell (SnO2–SiO x ) structure should result from the phase separation energy. The vertical growth and horizontal growth models may be controlled by adjusting the mass transfer coefficient. Free-standing nanorods only appear under the condition favorable to the vertical growth model. Shrimp-like, flower-like, and worm-like morphologies occur under the condition for the horizontal growth model.