Introduction

Carbon nanotubes (CNTs) and nanostructured metal oxide have been extensively investigated in the past two decades due to their many extraordinary physicochemical properties in industrial applications, such as photocatalysis and energy storage [1,2,3,4]. In most of these applications, efforts have also been devoted to fabricating these two classes of materials into CNT/metal oxide nanocomposites, aiming at a synergistic integration of their intrinsic properties in the new hybrid materials and thus improving the performance to meet new requirements imposed by advanced applications such as catalysis and nanotechnology [5,6,7,8,9,10,11,12,13]. The role of CNTs in these nanocomposites to form a 3D matrix is well acknowledged, which results in superior physical and chemical properties. In most of these nanocomposites, cable-like sheathed nanostructures have been grown, where CNTs serve as a core support for the metal oxide while the latter usually either filled the cavity and/or forms an overlayer. Meanwhile, the metal oxide components are mostly in the form of nanoparticles [6]. It is well known that metal oxides show shape-/size-dependent properties. Consequently, to greatly enhance the physical and chemical properties of metal oxide/CNTs nanocomposites, synthesis of heterogeneous nanostructured metal oxide and CNTs nanocomposites has long been a subject of interest.

To date, various approaches have been reported to obtain the heterogeneous metal oxides and CNTs nanostructures. A considerable number of chemical methods especially the template method have been employed for the generation of coaxial core–shell metal oxide nanotubes/CNTs nanostructures owing to their novel structures and outstanding properties, making them very promising for the next generation of energy storage applications [7, 14,15,16]. For example, Satishkumaar et al. have reported a template method involving using CNTs as a template for the synthesis of versatile oxide nanotubes, like SiO2, Al2O3, and V2O5 nanotubes [17]. In contrast, up till now, few reports on the synthesis of MONs/CNTS composites are available. Novel metal oxide nanosheets (MONs) on CNTs backbone have also sparked tremendous interest owing to their unique hierarchical porous structure, resulting in excellent physicochemical properties [6]. Moreover, compared with metal oxide nanoparticles/CNTs or metal oxide nanotubes/CNTs, MONs/CNTs nanocomposites manifest comparable or even better performance in certain applications. Hence, continuous efforts have been directed toward the synthesis of MONs/CNTs in the last few years. Recently, Ding et al. reported that the growth of SnO2 and TiO2 nanosheets with a small lateral size of ca. 100 nm upright on the CNT backbone could be achieved by facile solution methods [6]. Nevertheless, solution methods typically involve multiple steps of washing or purification that exclude them from wide application. Zhang et al. achieved the goal of preparing manganese oxide nanoflowers on vertically aligned carbon nanotube arrays (CNTA) by a combination of CNTA as the substrate and the electrodeposition technique [18]. The electrochemical performance for such binder-free manganese oxide/CNTA composite electrode exhibit superior rate performance. Although CNTA shows superiority over electrodepositing metal oxides due to their regular pore structure, prominent conductivity, and high surface area, the high manufacturing costs of CNTA could seriously hamper its use in industrial application. As mentioned above, all of the methods require multi-step synthetic processes, where the first step is generally to fabricate CNTs as the substrates, followed by the growth of metal oxides nanostructures. Therefore, innovative approaches for one-step and low-cost synthesis of MONs/CNTs composites would be quite advantageous to meet the need of future industrial production.

Here we report for the first time the synthesis of a 3D hierarchical porous structure consisting of MONs and carbon nanotubes grown simultaneously on stainless steel mesh by a one-step water-assisted chemical vapor deposition (CVD) process. Transverse and longitudinal cross-linked CNTs on MONs can be clearly observed when the synthetic time is prolonged longer than 5 min. Furthermore, a probable growth mechanism and the growth scenario are presented.

Experimental details

Simultaneous growth of heterogeneous and hierarchical MONs/CNTs nanocomposites on stainless steel

The experiments were carried out at the atmosphere pressure using a chemical vapor deposition (CVD) method, with H2 and Ar as the reaction carrier gas, methane as the carbon precursor, ferrocene (0.05 g) as the catalyst. The wall temperature of the quartz tube is measured through a pair of K-type thermal couples mounted in the central part of the oven, where there lie the raw stainless steel meshes. The CVD growth process occurs in the reactor center part. The schematic of the CVD system is shown in Fig. 1.

Figure 1
figure 1

a Schematic illustration of trace-water-assisted CVD system for the generation of MONs/CNTs hierarchical nanostructures. b Detailed demonstration of the CNTs/MONs growth process

Full size image

Stainless steel (AISI 304) meshes were used for this study, whose chemical composition is: Fe—77 wt%; Cr—20 wt%; Mn—2 wt%, other additives ~ 1 wt%. Firstly, stainless steel substrates were ultrasonically cleaned with ethanol and acetone for 10 min, and then rinsed thoroughly with deionized water, followed by drying with nitrogen. Subsequently, the inside walls of a horizontal quartz tube reactor were evenly wiped with cotton cloth soaked in deionized water. Then, the meshes supported by the quartz boat were placed at the center of the quartz tube passing through a tube furnace. The reactor was heated to a desired temperature up to 1020–1100 °C at a ramp rate of 25.5–27.5 °C/min under an ambient flow rate of Ar at 30 standard cubic centimeters per minute (sccm). Once the reaction temperature was attained, the quartz tube was moved. This stage is called the constant temperature calcination stage, which can be divided into stage 1 and 2, respectively. Stage 1 means that the initial growth of MONs and oxidation of the decomposed ferrocene, while stage 2 refers to the simultaneous growth of MONs/CNTs composites. In stage 1, both the water and ferrocene lying in the gas inlet direction in the area out of the quartz tube heating zone could be annealed at high temperature. The resulting steam serving as the oxidizer gas and the vaporized catalyst precursor were carried into the heated zone in a mixture gases flow composed of H2 and Ar with a flow rate of 30 sccm and 10 sccm, respectively. The flow rate of the mixed H2/Ar gas flow has been optimized based on the lateral size, quality, and quantity of the as-synthesized MONs in our previous reports [19]. Stage 1 was carried out for 1 min in all of our experiments. Subsequently, stage 2 began when methane with a flow rate of 90 sccm was introduced to the system while the H2/Ar flow rate was maintained unchanged, and the growth of MONs/CNTs nanostructures in stage 2 was carried out from 1 min to 1 h. It should be noted that the growth time of 1 min–1 h mentioned throughout the paper mainly refers to stage 2 in the constant annealing period. After carefully cooling the furnace to ambient temperature in H2 atmosphere, the MONs/CNTs composites were sampled.

Materials characterization

All of the scanning electron microscopy (SEM) experiments were performed by directly fetching as-grown samples without any subsequent processing using a Hitachi S-4800 instrument operating at 20 kV.

High-resolution transmission electron microscopy (HRTEM) images were recorded using a JEM-2100 electron microscope (JEOL) with an acceleration voltage of 200 kV. Elemental analysis of the specimens was performed with energy-dispersive system (EDS) attached to the transmission electron microscope. HRTEM imaging and EDS analysis were conducted by taking the commonly used method, where the as-prepared MONs/CNTs composites were detached from the stainless steel meshes by sonicating the stainless steel meshes in a neat ethanol solution for 15 min, and then a dispersion of nanoparticles in methanol was drop-cast on a carbon coated holey copper micro-grid, followed by a natural air-drying process.

Surface chemical characterization of the stainless steel meshes with the as-grown MONs/CNTs composites was carried out by X-ray photoelectron spectroscopy (XPS) using a spectrometer with Al Kα radiation (PHI 5700). The binding energy was calibrated with the C 1s position of contaminant carbon in the vacuum chamber of the XPS instrument. All of the specimens without any post-treatment were fetched directly for characterization in the corresponding XPS and Raman apparatus.

Raman spectra of meshes were collected by employing a Horiba Evolution instrument using a 632.8 nm laser as the excitation source.

Results and discussion

Our strategy to fabricate the hierarchical nanostructures essentially consisting of two classes of functional nanomaterials is briefly described as in Fig. 1. The detailed synthesis process can be found in the experimental section. The goal of facile and massive synthesis of such heterogeneous MONs/CNTs composites simultaneously on meshes could be achieved by one-step trace-water-assisted CVD method. Here, we study the effect of different experimental parameters on the synthesis of MONs/CNTs composites, where the concerns are mainly on the effect of synthetic temperature and time on the yield and morphology of MONs/CNTs composites. Additionally, the quality of the as-obtained CNTs is evaluated by Raman spectra. In this case, a probable growth mechanism is discussed in this present work.

Effect of temperature on the synthesis of MONs/CNTs composites

To understand and control the synthesis of such composites, we first examine the relationship between the synthetic temperature and the composites growth. In Fig. 2, SEM images give general structural information of the MONs/CNTs nanocomposites grown in the temperature range of 1020–1100 °C for 5 min. When the furnace temperature is varied from 1020 to 1100 °C, the mean diameter and abundance of CNTs are increased. The increased conversion of methane and yield of CNTs can be attributable to the enhanced thermal energy that plays a conspicuous role in decomposing even more ferrocene to small Fe nanoparticles. It has been summarized that higher temperature promotes hydrocarbons decomposition leading to more carbon generation and hence more tube and wall formation [20], which is in good agreement with our SEM observations. Surprisingly, the nanosheets standing vertically aligned on the stainless steel meshes can be identified. As depicted in Fig. 2a, semi-transparent and sparse MONs with an average lateral size less than 500 nm are obtained when the reaction temperature is elevated to 1020 °C. It is possible that the temperature of 1020 °C is not favorable for the large-area growth of MONs, and thus only a few MONs are identified to be intertwined with CNTs in diameter of less than 25 nm. Figure 2b illustrates that the whole morphology of the composites at 1060 °C is very similar to but not identical to that of as-grown MONs/CNTs at 1020 °C, since the mean lateral size of MONs remains approximately unchanged, but the mean diameter of CNTs is significantly increased to ~ 40 nm. Apart from that, the most abundant and densely grown CNTs at 1100 °C, shown in Fig. 2c and d, are of approximately the same size in diameter as compared to those grown at 1060 °C, and the one CNT cluster is easily observed in Fig. 2c. Moreover, when comparing Fig. 2d with Fig. 2a and b, it can be clearly seen that the mean lateral size of MONs is largely enlarged to around 1.2 μm, suggesting that 1100 °C (the maximum temperature in our furnace) is the optimal synthetic temperature for boosting the synthesis of both MONs and CNTs to the next level and self-assembly of the hierarchical nanocomposites in our case.

Figure 2
figure 2

SEM images of MONs/CNTs nanostructures grown at a 1020, b 1060 and c, d 1100 °C for 5 min each. The width of the MONs increases with temperature. The optimal synthetic temperature in our experiments is 1100 °C

Full size image

Dependence of MONs/CNTs nanocomposites morphology on the synthetic time

To better understand the growth process of MONs/CNTs nanomaterials, we have examined the effect of synthetic time in stage 2 on the fabrication of MONs/CNTs composites. Here, the synthetic time does not contain the time for stage 1 (see more details in the experimental part). For the sake of specificity, the growth temperature was set to 1100 °C, and the other relevant conditions were also kept constant. SEM imaging of the MONs/CNTs grown for 1 min (Fig. 3a) discloses that there are few straight CNTs growing on the mesh and one can observe the nanosheet-like structures with nanoparticles standing both on the stainless steel and nanosheets surface. The generation of some straight CNTs may due to the catalytic decomposition of methane by Fe nanoparticles, presented in the inset of Fig. 3a. Meanwhile, these nanoparticles are supposed to be predominantly composed of iron oxides (FexOy) and Fe nanoparticles when the synthetic time is not enough long, the formation of FexOy gradually occurs when Fe nanoparticles arising from the decomposition of ferrocene meet and subsequently may be easily oxidized by heated water vapor. Hydrogen may partially reduce the iron oxides. Thus iron oxides with different oxidation states may form. Specifically, the iron oxides catalyst nanoparticles attach directly to the outside surface of MONs, resembling shells decorated with pearls. It can be presumably deduced that the growth direction of MONs is nearly perpendicular to the direction of the reaction gas flow. When the growth process is extended to 5 min, the phenomenon of catalyst nanoparticles adhering to the nanosheets surface is also identified, illustrated in Fig. 3b, c. In the meantime, it can be observed that those transverse and longitudinal cross-linked CNTs with diameters of between 10 and 112 nm grow everywhere and the abundance of CNTs and MONs is found to be substantially improved as compared to Fig. 3a. This is mainly due to the fact that the catalysts have enough time to catalyze the pyrolysis of methane, and the spatial anisotropy distribution of FexOy nanoparticles favors the growth of curved and entangled CNTs in any direction. When the catalysts adhere to the surface of MONs, the catalytic decomposition of carbon precursor can take place on the surface of MONs, thus the as-grown CNTs are attached to the MONs facet. Therefore, CNTs are closely intertwined with MONs. Figure 3d depicts that several slices of MONs aggregate together when the growth time is prolonged to 10 min, which is an indirect manifestation of the dense growth of MONs. Figure 3e, f demonstrates the typical helical coiled CNTs (indicated by red arrows) can be synthesized with an extended reaction time of 1 h. It is obvious that the helical coiled CNTs are generally made up of two separate CNT with nearly equal diameter of about 25 nm that spirally and symmetrically curved together.

Figure 3
figure 3

SEM images of MONs/CNTs nanostructures grown at 1100 °C for a 1 min, b and c 5 min, d 10 min, e, f 1 h. The inset of a shows few typical straight CNTs grown at 1100 °C for 1 min. c and f are the enlarged views of the selected region of (b) and (e), respectively

Full size image

Evidence for MONs/CNTs composites from TEM analysis

The detailed morphologies and microstructures of the MONs/CNTs composites have been further examined by TEM. The representative TEM image of prepared MONs/CNTs composites displayed in Fig. 4a feature the MONs entangled with CNTs. Figure 4a also shows that the MONs display the metallic luster. Figure 4b shows that CNT has a strong attachment to MONs and provides evidence to support their strong bonding instead of merely surface contact. This can be attributed to FexOy nanoparticles strongly attached to the MONs, which serves as a key factor in achieving the excellent connection of CNTs with MONs, where the high temperature and long annealing time also play a vital role. The EDS measurement was taken from the selected area of Fig. 4b, indicated by circle uncovered in Fig. 4d. The EDS results demonstrate that the MONs are composed mainly of Fe/Cr/Mn oxides. The content with respect to carbon derives from CNTs. As depicted in Fig. 4c, HRTEM characterization of samples grown at 1100 °C for 5 min demonstrates that the typical fish-bone-type CNTs with an outer diameter of about 25 nm and a wall thickness of about 6 nm have many carbon sheets with a distorted angle (θ: ~ 30°). In contrast, it is well know that the straight tubes commonly exhibit a cylindrical structure where the graphite sheets are aligned parallel to the axial direction. Meanwhile, the fish-bone-type CNTs with graphite layer spacing ~ 0.34 nm (corresponding to the spacing of the graphite layer (002) crystal surface) are seen in the presence of open ends. This phenomenon is confirmed by SEM images shown in Fig. 4e, f. Previous reports have shown that in the absence of water, high-purity straight CNTs with closed ends could be fabricated [21,22,23]. Lim et al. proposed that the tip of CNTs was more likely to be oxidized since it was exposed to air or other oxidizer, and then the metal catalysts on the tip of CNTs could be oxidized, expanded, and removed easily [24]. Similarly, it is possible that the formation of open-ended CNTs can be largely due to the constant oxidation of iron nanoparticles by water in our case.

Figure 4
figure 4

a and b The typical TEM images of MONs/CNTs hierarchical nanostructures. c HRTEM images of fish-bone type CNTs grown at 1100 °C for 5 min. f EDS analysis of the CNTs and MONs composites, the spectrum was taken from the marked area shown in (b). e SEM and f further enlarged SEM images of MONs/CNTs nanostructures grown at 1100 °C for 5 min

Full size image

XPS analysis

As shown in Fig. 5, the composition and oxidation state of elements on the selective samples as oxidized mesh at 1100 °C for 5 min were examined by XPS survey scan. The wide XPS spectrum (Fig. 5a) shows the peaks attributed to the core levels of C 1s, O 1s, Cr 2p, Fe 2p, and Mn 2p. Figure 5b presents the Cr core level XPS spectra of the oxidized mesh. The Cr 2p spectrum can be deconvoluted into three pairs of doublets and a satellite peak, with the doublets at 588.1/579.1 eV in agreement with CrO3 and the doublets at 586.9/577.8 and 586.4/576.7 eV corresponding to CrOOH and Cr2O3, respectively [25,26,27]. Appearance of satellite peak at 582.5 eV can be attributed to the Cr metal in the substrate [28]. Figure 5c shows that Fe 2p spectrum can be deconvoluted into two pairs of doublets and a satellite peak, with the doublets at 726.3/710.7 and 723.6/713.6 eV corresponding to Fe3+ and Fe2+ [29], respectively. The peak at 718.2 eV is a satellite peak for the above four peaks, representing the coexisting of Fe3+ and Fe2+ [30]. The Mn 2p spectrum shown in Fig. 5d can be deconvoluted into three pairs of doublets and a satellite peak, with the doublets at 654.6/642.9, 653.8/642, and 652.4/640.8 eV corresponding to Mn4+, Mn3+, and Mn2+, respectively [31, 32]. The presence of satellite peak at 648 eV can be attributed to the Mn2+ in the substrate [33]. The fitting results demonstrate that water has oxidized the stainless steel mesh, resulting in Fe/Cr/Mn oxides with different oxidation states.

Figure 5
figure 5

a XPS wide-scan spectra; high-resolution XPS spectra of b Cr 2p, c Fe 2p and d Mn 2p of stainless steel annealed at 1100 °C for 1 h

Full size image

Raman analysis

To further evaluate the crystalline feature of the as-grown CNTs in the water-assisted CVD process, normalized Raman spectra of the MONs/CNTs composites at different growth time were recorded as shown in Fig. 6. Spectrum A was taken from an area rich in CNTs, while spectrum C was taken from the regions of MONs/CNTs composites. The remarkable difference in the CNTs quality is supported by the data extracted from Raman spectra. As can be seen, both the samples grown at 1100 °C for 5 min and 1 h exhibit a disorder band (D-band) at ~ 1328 cm−1. The distinct differences as seen in the Raman spectra are as follows. The as-grown 5-min CNTs show a sharp tangential graphite band (G-band) at 1580 cm−1. It is well known that the G peak originates from the tangential vibration of sp2 carbon networks [34], and the D-peak is related to the breathing modes of six-atom rings and requires a defect for its activation, i.e., the peak arises from lattice disorder [35]. This implies that the CNTs grown in the tubular reactor for 5 min show a high overall degree of graphitization. Comparison between spectra A and B shows that the introduction of trace water enhances CNTs graphitization. Yet, the as-grown 1-h CNTs show an asymmetrical wide peak appearing at 1596 cm−1 with an intensity less than D-peak, confirming the poor graphitization of CNTs grown, as shown in the spectrum C. Recent theoretical simulations reveal that trace water can react with the dangling bond on carbon, thus inhibiting the formation of amorphous species [36]. It was pointed out that the synthesis of CNTs for too long or too short times is not beneficial to the quality of carbon nanotubes obtained [20]. Xie et al. proposed that when the synthetic process is implemented for a long enough time, the non-catalysis of pyrolytic carbon and surface etching through carbon gasification are competitive during the CNT synthesis [37]. They assigned this period to the passive growth stage, where the surface etching through gasification becomes the dominant process, causing increased surface defects of CNTs. Therefore, it is possible that when the reaction time is extended to 1 h, the water on the tube walls of the reactor is totally gone. Simultaneously, the nearly no-catalysis growth of CNTs is occurring. As a consequence, low-quality carbon may deposit on the outer wall of CNTs. Surface etching may become the dominant process, where H2 serves as the etchant.

Figure 6
figure 6

Raman spectra of CNTs grown at 1100 °C for (a) 5 min and (c) 1 h by trace-water-assisted CVD. b Raman spectrum of CNTs grown at 1100 °C for 5 min without the introduction of water to the CVD system

Full size image

Note that spectrum C also exhibits some other peaks. The Raman peaks appearing at 296, 349, and 555 cm−1 are related to Cr2O3 [38], whereas the peaks seen at 619 and 681 cm−1 are related to Fe3O4 or Mn3O4. Here, it should be pointed out that peaks inherent to Mn3O4 lattice are very close to those of Fe3O4, and thus it is not possible to accurately assign these peaks [39, 40]. It is consistent with our previous report that the MONs grown by trace-water-assisted high-temperature annealing process are mainly composed of Cr2O3, followed by (Fe, Mn)3O4 [19].

Growth mechanism of MONs/fish-bone-type helical CNTs composites

The above microscopic evidence suggests a probable hierarchical formation scenario and a consequent growth mechanism. Figure 7a reveals that the growth of nanoscopic hierarchical MONs/CNTs structures in the one-step CVD growth process can be deemed as a two-stage process. Initially, the water lying on the wall of quartz tube is evaporated to steam and subsequently carried to the mesh by the flowing gas in stage 1, the surface defects or cracks in the stainless steel substrate serves as the nucleation sites and MONs initiate growth at these nucleation sites. The formation of MONs follows a tip growth mechanism as proposed in our previous reports [19]. As a consequence of the large specific surface area of MONs, the decomposed Fe nanoparticles deposit on the surface of the nanosheets. Also, Fe nanoparticles can deposit on the mesh surface, Fe nanoparticles are oxidized into FexOy when they meet the water vapor. Iron oxides nanoparticles then catalyze the decomposing of methane to carbon atoms and thus CNTs directly grow on the surface of MONs and stainless steel meshes. In stage 2, the typical fish-bone-type helical CNTs are observed. The probable growth mechanism of CNTs grown in stage 2 for 1 min to 1 h is briefly shown in Fig. 7b. It has been well reported that employing the FexOy catalytic nanoparticles to decompose carbon precursor could successfully form fish-bone-type CNTs [41]. So in our case, it is possible that the as-formed FexOy nanoparticles may be the key factor for the synthesis of fish-bone-type CNTs. Sawant et al. proposed that structural and composition nature of catalyst do play an important role in the formation of fish-bone-type CNTs [42]. Meng et al. put forward that the growth of helical CNTs in their trace-water-assisted CVD system can be explained that trace amount of water favors the dramatic anisotropy of the catalyst surface [36]. Accordingly, carbon atoms formed by the decomposition of the carbon precursor on the surface of the catalyst may be separated in different directions through the crystal face, and thus forming the helical CNTs (Fig. 7b). However, the exact diffusion process at the atomic level is still unclear and needs to be further investigated. The size of FexOy nanoparticles is supposed to keep getting larger, and the structure of FexOy nanoparticles also varies with the increase in synthetic time. Thus, decomposed carbon atoms could form graphite structure through different FexOy crystal facets. Different generation rates of the graphite structure along the various directions of the catalyst surface feature a pronounced preference to form the helical structure. An earlier report demonstrated the kinetically controlled growth of helical and zigzag shapes of carbon nanotubes [43]. Gao et al. reported that Al2O3 was put into the Fe(NO3)3 solution and drying the solution via rotary evaporation and baking, then the obtained catalysts were grounded into powder, and the hydrogen-reduced catalysts containing Fe, FeO, and Fe2O3. Thus, both helical and zigzag CNTs were synthesized by catalytic decomposition of ethylene. The growth model proposed by Gao et al. and other groups claimed the essence that the pairing of pentagonal–heptagonal (P–H) carbon rings played in generating the helical CNTs [44, 45]. Ihara and Itoh suggested that a helical tube can be grown if the orientations of the P–H pairs rotate along the body of the tube [46]. Therefore, it is possible that the graphite structure generated on the surface of FexOy may exist not only in the commonly hexagon rings, but also with various P–H pairs, and thus curved and helical CNTs grow in a self-assembled fashion.

Figure 7
figure 7

a Schematic illustration of the growth mechanism for the one-step water-assisted CVD growth process of MONs/CNTs composites. b Probable growth mechanism of straight and helical CNT in the CVD process

Full size image

Compared with other methods reported before, where commonly involved the multiple steps to manufacture the MONs/CNTs composites, our process mitigates the need of CNT templates, which is simple and feasible. The research of MONs/CNTs composites is still a rather new scientific field, and the exact growth mechanism needs to be further demonstrated. But it is clear that the simultaneous one-step growth process presented above could pave a new way to fabrication of hierarchical MONs/CNTs composites.

Conclusion

In summary, a novel hierarchical MONs/CNTs nanostructure has been experimentally synthesized by a one-step CVD method. The size and morphology of MONs and CNTs can be tuned by varying the synthetic time and the annealing temperature. The optimal synthetic temperature for boosting the synthesis of the hierarchical nanocomposites is 1100 °C, which is the maximum temperature in our furnace. The water vapor inside the reactor is totally gone and the nearly no-catalysis growth of CNTs occurs when the reaction time is extended to more than 1 h. The generation of MONs/CNTs composites follows a two-stage process, where stage 1 includes growth of MONs and the oxidation of decomposed ferrocene nanoparticles to form FexOy. And the formation of FexOy favors the growth of helical CNTs in stage 2, and trace water may be the key factor for the dramatic anisotropy of the obtained FexOy surface, leading to the different segregation rate for carbon atoms along the catalyst surface and the formation of graphite structures accordingly. The simultaneous growth of MONs and CNTs composites and the strong adhesion between CNTs and MONs may boost these new hierarchical nanocomposites to be a fertile ground for future research.