Single-step synthesis of carbon nanotubes/iron/iron oxide composite films through inert-ambient CVD using ferric acetylacetonate as a precursor

Abstract

We have developed a unique single-step chemical vapor deposition (CVD) route for the synthesis of composite thin films containing carbon nanotubes (CNTs). CVD was carried out in an inert ambient using only iron(III) acetylacetonate as the precursor. Depositions were conducted at 700 °C on stainless steel substrates in argon ambient in the absence of any reactive gases (such as oxygen, hydrogen). By changing the deposition parameters, especially the pressure in the CVD reactor, the form of carbon deposited could be changed from amorphous to carbon nanotubes, the latter resulting in Fe–Fe₃O₄–CNT films. X-ray diffraction, Raman spectroscopy, X-ray photoelectron spectroscopy, and electron microscopy together confirm the formation of the three-component composite and illustrate the nanoscale mixing of the components. Elemental iron formed in this process was protected from oxidation by the co-deposited carbon surrounding it. Irrespective of the substrate used, a composite coating with CNTs was formed under optimum conditions, as verified by analyses of the film formed on polycrystalline alumina and silicon substrates.

---

Introduction

Metal oxide–carbon composites offer the opportunity to explore their potentially superior material characteristics relative to the constituents, as the properties of individual components can result in a synergetic effect when they are parts of a composite. The many desirable attributes of carbon – such as its mechanical strength, stiffness, electrical conductivity, ready availability, cost-effectiveness, low density, chemical stability, and amenability to processing for tailoring its form and properties – make “carbons” sought-after candidates in composite formation. Composites of metals, and metal oxides with various forms of carbon such as nanotubes, fibers, spheres, aerogel, graphene etc., have been prepared and studied for their applications in sensors, catalysis, energy storage devices and structural materials, giving rise to improved functional materials.^1–8 In particular, carbon nanotubes (CNT) and their composites have been the subject of much worldwide research and development (R&D) effort over the last two decades.^5,7,9–11

Most of the reported methods for preparing CNT–metal oxide composites are multi-step processes using distinct starting materials (precursors) as sources for carbon and the metal oxide. Solution-based methods involve separate synthesis of CNTs, followed by adding them to the solution containing the precursor to the metal oxide. From this solution, the metal oxide is obtained through the sol–gel method, chemical precipitation, impregnation, solvothermal methods or electrodeposition, in the presence of CNTs, thus forming a composite.^2,9,12–17 The primarily two-step procedure may include an additional step where CNTs are acid-treated for better dispersion in the solution.^15,18

Thin films of metal oxide/CNT composites are usually obtained through distinct deposition steps in a preset sequence using appropriate starting materials for each of the constituents. Typically, CNTs are obtained by chemical vapor deposition (CVD) and the oxide phase deposited by sputtering, CVD or electrodeposition.^16,19–22 In contrast with the aforementioned methods, our approach has been to use a metalorganic compound (in this case ferric acetylacetonate), which contains direct Fe–O bonds as well as hydrocarbon groups in its molecular structure, as precursor in CVD and to arrive at the deposition conditions needed to obtain metal oxide–CNT composite thin films.

The present effort was prompted by the results obtained earlier in the author's research group on MnO/C composites obtained by a similar inert-ambient CVD process employing manganese acetylacetonate as precursor. The MnO/C coatings on stainless steel proved to be excellent electrodes for thin film supercapacitors.²³ Magnetic measurements made on such coatings (on alumina substrate) have been reported.²⁴ Our work on inert-ambient CVD using Fe(acac)₃ was initiated with the expectation that similarly excellent (thin film) electrode materials can be obtained. What we obtained instead were the three-component CNT/Fe/Fe₃O₄ composites reported here.

Iron exists in +3 and +2 oxidation states, forming several distinct oxide phases, of which α-Fe₂O₃ is the most stable. Fe₃O₄ and γ-Fe₂O₃ are technologically important due to their magnetic ordering, the latter finding applications in magnetic recording media.^25–27 Even though iron oxides are very well known materials, they are today being explored in their nanometric form for their magnetic, catalytic, sensing, and energy storage properties.^27–30 Nanoparticles of iron oxides and their nanostructured thin films have been prepared by various chemical methods. In particular, CVD of iron oxide films has been carried out in flowing oxygen, using solid precursors such as ferric acetylacetonate [Fe(acac)₃] or iron carbonyl [Fe(CO)₅], with argon as the carrier gas, typically in the temperature range of 350–550 °C.^31–34 Careful control of the oxygen partial pressure and substrate temperature is required in such a process to obtain Fe₃O₄ or the metastable phase γ-Fe₂O₃ (rather than α-Fe₂O₃). Furthermore, CVD is widely employed to obtain CNTs, using various hydrocarbons, alcohols, and organometallics, such as metallocenes of Fe, Ni and Co, as precursor.^35–38

We report here on a single-step process for the formation of a thin film composite of CNTs, iron, and iron oxide by CVD. This method uses a metalorganic complex of iron as the single-source precursor, which contains both iron–oxygen bonds and hydrocarbon moieties in its molecular structure. Specifically, the CVD process was carried out in an inert ambient using ferric acetylacetonate as the precursor. Under certain deposition conditions, in the absence of any reactive gas such as O₂, H₂, etc., a three-component composite film consisting of elemental iron, Fe₃O₄, and carbon nanotube is obtained. By “tuning” the deposition conditions, the nature of the carbon in the composite can be varied from amorphous carbon to CNTs. We have recently provided a brief report rationalising the formation of Fe/Fe₃O₄/CNT composites through equilibrium thermodynamic modeling of the inert-ambient CVD process.³⁹

It is relevant to note here that C/Fe/Fe₃O₄ composites, including those with carbon in the form of CNTs, have been studied by various investigators, with different applications in mind. These include: as anode material in lithium-ion batteries;^40,41 in waste water treatment;⁴² in catalysis and the abatement of the environmental hazards.^43,44

Experimental

The molecular structure of the precursor, ferric acetylacetonate, Fe(acac)₃, is shown in the inset of Fig. 1. The complex is a crystalline solid that sublimes above ∼140 °C. Fig. 1 shows data from the simultaneous thermogravimetry (TG) and differential thermal analysis (DTA) of Fe(acac)₃ (obtained from a SDT Q600 TG/DT Analyzer, TA Instruments, USA). TG data show that, as temperature is raised, a modest weight loss begins at about 125 °C. Much of the weight loss occurs in a single step monotonic loss in weight as the temperature is raised. Weight loss occurs in a single step spanning ∼170–220 °C, leaving a substantial residue. The shaded area in the TGA graph indicates the range of temperatures, ∼150–160 °C, at which the complex may be maintained so that an adequate (and optimum) flux of the precursor vapor is available during the CVD process. Over this range, the precursor flux would be a mild function of temperature. It is to be noted from the DTA graph that the complex melts or decomposes only at temperatures significantly higher than the vaporization temperature employed in the present work.

Fig. 1 Simultaneous TG-DTA data for Fe(acac)₃ obtained under a flow of dry nitrogen, at a heating rate of 10 °C min⁻¹.

Depositions were carried out in a horizontal, hot-wall, low-pressure CVD reactor built in house. The reactor chamber is a fed quartz tube wherein cleaned substrates were placed. The precursor, ferric acetylacetonate (Alfa Aesar, 99.9%), was taken in a quartz boat in finely powdered form, and placed inside the vaporizer. Ultra-high purity argon served as both the carrier and purging gas. Intentionally, no reactant gas, such as oxygen or hydrogen, was used in the CVD process. Calibrated electronic mass flow controllers (MFC, MKS-made, USA) were used for the precise control of gas flow. The CVD chamber was evacuated by a rotary pump and the pressure in the reactor is measured by a heated capacitance manometer (Baratron, MKS-made, USA) placed close to the reaction chamber. Pressure in the chamber was adjusted using a precision needle valve in line with the throttle valve. Depositions were carried out mostly on stainless steel (SS316) substrates and in some cases on single-crystalline silicon [Si(100)] and polycrystalline alumina. Depositions were conducted at 700 °C, with carrier gas (argon) flow rate in the range 20–40 sccm, and the deposition pressure in the range 5–30 torr.

Characterization

X-ray diffraction (XRD) patterns of films were recorded using a PanAnalytical powder diffractometer with Cu-K_α radiation (λ = 1.5418 Å). Scanning electron microscopy (SEM) data were obtained in a field emission scanning electron microscope (FESEM, Zeiss Ultra55). Transmission electron microscopy (TEM) was carried out on film samples separated carefully from the substrate (TECNAI F30). X-ray photoelectron spectroscopy (XPS) was performed using Mg-K_α radiation (1253.5 eV) in a high-resolution spectrometer equipped with auto charge neutralization (AXIS ULTRA, Krotos, UK). Raman spectra were obtained using an argon ion laser of wavelength 514 nm (Horiba Jobin-Yvon LabRAM HR 100 spectrometer).

Results and discussion

All the depositions resulted in black, adherent coatings, suggesting the presence of elemental carbon in them.

The morphology of the coating (Fig. 2(a)) obtained at 5 torr shows the aggregation of submicron-sized, faceted particles which appear to be crystalline. The coating was examined for the presence of elemental carbon by Raman spectroscopy, which is a powerful tool for affirming the presence of iron oxides as well as the different forms of carbon. The Raman spectrum of the film (Fig. 2(b)) shows peaks centered at 1340 cm⁻¹ (D), 1590 cm⁻¹ (G), and 2700 cm⁻¹ (G′), each of which is characteristic of carbon, thus confirming the presence of elemental carbon. The G peak is characteristic of sp² carbon, representing the E_2g vibrational mode of graphitic carbon. The D peak at 1350 cm⁻¹ is due to the A_1g lattice vibration mode of graphite, which arises only in the presence of defects, but is otherwise absent in graphite.^45,46

Fig. 2 (a) SEM micrograph and (b) Raman spectrum of the composite film on SS316 obtained at 5 torr.

The G′ band at 2700 cm⁻¹ is a second order feature that appears in the presence of structural ordering, indicating a well-crystallized graphite lattice. For the present sample, this peak is of a relatively low intensity, with slight broadening. The broad and overlapping D and G peaks indicate the presence of a substantial proportion of amorphous carbon in the coating.^47–49 The Raman spectrum also features a peak at 667 cm⁻¹, which is due the A_1g vibrational mode of magnetite (Fe₃O₄); weaker peaks corresponding to T_2g vibrational mode of Fe₃O₄ are also seen, at 520 and 305 cm⁻¹.^50–52 Raman analysis therefore confirms the formation of a composite film comprising carbon and iron oxide (Fe₃O₄).

An increase in the CVD process pressure to 10 torr resulted in the formation of copious amounts of carbon nanotubes (Fig. 3(a)), with a few large particles (akin to those formed at 5 torr) underneath the CNTs (Fig. 3(b)). Correspondingly, the Raman spectrum of these coatings (Fig. 3(c)) shows sharp and strong G and G′ peaks, along with the D peak. The absence of the radial breathing mode (RBM) implies the absence of single-walled carbon nanotubes (SWCNTs), thus indicating that the CNTs formed are multi-walled carbon nanotubes (MWCNTs).⁵³ The Raman spectra also feature a peak at 667 cm⁻¹, corresponding to magnetite. (To avoid the possibility of sample oxidation under laser impact during Raman analysis, laser power was kept <3 mW).

Fig. 3 (a), (b) SEM micrograph of film obtained at 10 torr displaying CNTs and underlying particles (c) Raman spectrum of such a film, which confirms the presence of magnetite and CNTs.

Likewise, as depositions were conducted at increasing pressures, the formation of a composite film of CNTs and iron oxide was observed upto 30 torr. The surface morphology of the composite film obtained at 30 torr and the corresponding Raman spectrum, confirming its composite character are shown Fig. 4(a) and (b). The relative intensities of the G and D peaks indicate the presence of crystalline carbon with a relatively low density of defects.

Fig. 4 (a) SEM micrograph and (b) Raman spectrum of the composite film on SS316 obtained at 30 torr.

These data show that, under the CVD conditions cited, the formation (deposition) of a CNT–Fe₃O₄ composite takes place through the simultaneous deposition and intergrowth of CNTs and Fe₃O₄, so that the two components are “mixed' on a very fine scale. This is reflected clearly in the SEM micrograph of the film shown in Fig. 4(a), 5(a) and in the bright-field TEM image Fig. 5(b).

Fig. 5 (a) Cross-sectional SEM of the CNT/magnetite composite deposited on SS316 @30 torr (b) TEM bright-field image of CNT/iron oxide composite with (c) showing the filled inner cavities of CNTs.

The TEM micrograph (Fig. 5(b)) shows MWCNTs surrounded by particulate iron oxide. The length of CNTs formed is in the range l–10 μm, with the diameter in the range of 80–150 nm. (Some of the CNTs may be much longer than 10 μm, though this cannot be ascertained due to the tangled nature of growth.) The CNTs obtained have well-graphitized walls and many are found to be “filled”. That is, a short length of the tubes encloses a material of a high atomic number, presumably related to iron, as indicated by the TEM micrographs (Fig. 5(c)).

Thus, inert-ambient, low-pressure CVD with the metalorganic complex Fe(acac)₃ provides a single-step chemical process for the formation of a composite film of CNTs and iron oxide. By varying the reactor pressure, it was possible to vary the form of carbon obtained, from amorphous to crystalline CNTs.

Characterization of Fe/Fe₃O₄/CNT composite

The composite nature of the deposits was further corroborated by XRD. Fig. 6 shows the typical XRD pattern of the coating on SS316 formed at 30 torr and 700 °C. Diffraction peaks corresponding to magnetite (JCPDS file no. 19-0629) and the (002) reflection of graphite (JCPDS file no. 75-1621) were observed. Along with these two, there was a sharp peak at 44.7°, which could be due to either elemental iron or iron carbide. Similar XRD patterns were also observed on films deposited at lower pressure albeit with the absence of graphite (002) peak. Since deposition was carried out with a precursor containing hydrocarbon groups, which might result in a carbon-rich environment at a high temperature with a stainless steel substrate present, the formation of various iron carbides could not be ruled out.

Fig. 6 XRD pattern of the composite film revealing the different constituents of the composite coating formed on SS316 at 700 °C and 30 torr.

Therefore, films were deposited simultaneously on alumina and SS316 substrates. The peak at 44.7°, attributable to elemental iron, is present in the XRD pattern (Fig. 7) of the film grown on alumina substrate as well; this shows the formation of Fe is independent of the substrate used.

Fig. 7 Comparison of XRD patterns of alumina substrate before and after deposition at 700 °C. The standard pattern for α-Al₂O₃ is shown in blue and that of the coated substrate in black. The magnified pattern on the right shows evidence for the deposition of Fe.

XPS analysis of the composite coatings was carried out to confirm the presence of elemental iron and Fe₃O₄, The Fe 2p spectrum (Fig. 8) consists of photoelectron peaks corresponding to both Fe²⁺ (709.5 eV) and Fe³⁺ (711.1 eV), whereas the Fe³⁺ satellite feature at 719 eV was absent. These data, together, confirm the presence of Fe₃O₄.^54,55

Fig. 8 Fe 2p and C 1s spectra of the coating of the coating formed on SS316 at 30 torr.

A sharp photoelectron peak at 707.1 eV in the 2p_3/2 spectrum, corresponding to elemental iron, could be clearly discerned.^40,54–56 Thus, XPS data convincingly establish the presence of elemental iron in the film. Furthermore, the core level C1s spectrum does not show a peak at 283 eV, which proves the absence of iron carbide (Fig. 8).

TEM analysis was carried out to identify the constituents of the films on the nanoscale. The bright-field TEM image (Fig. 9(a)) shows relatively large (sub-micron sized) and well-faceted entities with a complex structure. The electron diffraction pattern taken on the faceted side yields a single-crystalline spot pattern corresponding to elemental iron, which is superposed on diffraction rings due to carbon and polycrystalline Fe₃O₄. This reveals convincingly the simultaneous formation and intimate coexistence of iron with iron oxide and carbon at the nanoscale.

Fig. 9 (a) Bright-field TEM image of a faceted particle present in the composite film formed at 10 torr (b) SAED pattern taken on particle confirms various phases which are part of single particle. Ring patterns due to iron oxide, carbon, and diffraction spots due to elemental iron could be observed (zone axis-[001]).

From XRD, XPS, and TEM measurements, it is apparent that low-pressure, inert-ambient CVD using Fe(acac)₃ results in the formation of coatings comprising iron, iron oxide (Fe₃O₄), and carbon.

A better understanding of the CVD growth process leading to the formation of composite coatings can be gained from TEM shown above (Fig. 9), together with a careful and detailed SEM examination of the coatings formed at different pressures in the CVD reactor. On steel substrates, the initial layers are found to be made of large (500–1000 nm) crystals of Fe and Fe₃O₄, along with carbon. Often, Fe and Fe₃O₄ are found to be parts of the same “grain”. Carbon is deposited on these entities and surrounds them.

There is thus evidence that two processes take place -

• Formation of large (500–1000 nm) particles, which consist of Fe, Fe₃O₄, and carbon (Fig. 10(a)).

Fig. 10 An illustration of the different film morphologies obtained on SS316 under different CVD conditions, using Fe(acac)₃ as the precursor.

• Formation of CNTs, along with iron oxide nanoparticles (Fig. 10(c)).

The formation of large particles appears to be the first step in the formation of the deposit, which may continue, resulting in morphology shown in Fig. 10(a). Alternatively, carbon nanotube growth can take place, depending on the deposition conditions, leading to the film shown in Fig. 10(c). In Fig. 10(b), there are large grains, which are few in number and are covered with tubular carbon structures. It is to be noted from Fig. 9 and 10 that the crystals of elemental iron formed in the process can be rather large, perhaps as large as several hundred nm. Given the relatively large size of the particles of iron that presumably catalyse the formations of CNTs, those that are formed are expected to be multi-walled; this is evidenced by Raman analysis.

CNT growth on other substrates

As noted earlier, depositions were also conducted at 700 °C simultaneously on SS316, Si(100), ceramic alumina (Al₂O₃) substrates, placing them side by side in the deposition chamber. CNTs were found to form also on Si (100) and polycrystalline alumina, as shown by SEM micrograph in Fig. 11 and further confirmed by Raman spectroscopy (Fig. 11(c)).

Fig. 11 SEM micrographs of the coating formed on (a) Si (b) Al₂O₃ at 30 torr, displaying CNT growth, with differing morphologies (c) Raman spectrum of the film formed on Si(100) at 700 °C and 30 torr.

The XPS spectrum of film on silicon also shows the presence of Fe and Fe₃O₄ (ESI†). Although CNTs are deposited on Si (100), there is no evidence in the Raman spectrum for the deposition of iron oxide, indicating that a much smaller proportion of Fe₃O₄ is perhaps formed on silicon than in the film on SS316. This illustrates qualitatively the influence of the chemical nature of the substrate on the pathways of chemical reactions in the CVD process.

The results presented above prove that CNT formation is an intrinsic part of the inert-ambient CVD process with Fe(acac)₃ as the precursor, and that it is independent of the substrate used. The proportion of CNTs and Fe₃O₄ formed, the structural characteristics of CNT, and thus the final morphology of the film, were found to be substrate-specific. For example, films obtained on silicon, SS316, and alumina, simultaneously, display different morphologies (Fig. 4(a), 11(a) and (b)).

Metal-organic route for composite formation

The results described above establish clearly that metal acetylacetonates in particular – and metal β-diketonates in general – can be used as single-source precursors for obtaining carbonaceous composites through inert-ambient CVD.²⁴ It has been generally believed that the formation of CNTs by CVD requires (or is facilitated by) the presence of fine particles of a transition metal to catalyse the high-temperature decomposition of a hydrocarbon or other carbon source. The formation of CNTs in the present process was unexpected – even remarkable – because the chemical source available was only Fe(acac)₃, a metalorganic complex in which Fe is bound to oxygen and carbon is bound to oxygen and hydrogen. A mechanism for the formation of such a composite may be formulated (as below) by considering results of the detailed characterization of the deposits carried out. The most significant feature of the depositions is the formation of finely divided elemental iron, which is ordinarily prone to rapid oxidation. The samples were analysed by XPS months after they were prepared, testifying to the stability of iron in it against oxidation. This stability is due to the “passivation” provided by the carbon that envelops the elemental iron – either CNTs enclosing it or amorphous carbon that surrounds it, evidenced by SEM and TEM analysis (Fig. 9).

The formation of elemental iron which, in turn, catalyses the formation of carbon nanotubes from the hydrocarbon moieties of the iron complex, may be explained as follows. At the elevated temperature and inert (highly reducing) ambient of the CVD chamber, the decomposition of the iron complex may be expected to lead to the formation of Fe_1−xO (wuestite), Fe₃O₄ (magnetite), and iron carbide, Fe₃C, and elemental carbon.³⁹ Iron carbide is unstable in a carbon-rich environment at high temperatures, and decomposes into Fe and C.^57–59 The formation of Fe catalyses the formation of CNTs. The magnetite in the composite is likely to be formed during the cool-down from the deposition temperature, due to the disproportionation of Fe_1−xO into Fe and Fe₃O₄ at temperatures <567 °C.⁶⁰ Thus, the product is a composite of Fe, Fe₃O₄, and C, with the carbon “turned into” the CNT form under optimum CVD conditions.³⁹

Since the deposition involves the formation of Fe⁰, the process becomes very sensitive to the presence of oxygen. Therefore, minute leaks in the CVD system can pose problems of reproducibility. Furthermore, any CVD condition that influences the decomposition pathways of the precursor affects the composition of the deposit – specifically, the formation of CNTs – significantly. As the difference in the free energy of formation of different forms of carbon is smal, This is probably the reason why, in the present effort, the formation of CNTs is found to occur over a limited range of reactor pressures.

Finally, comments are in order concerning some measurements made on the composite CNT/Fe/Fe₃O₄ coatings. The presence of a substantial proportion of carbon in the composite leads to appreciable electrical conductivity in the composite (measured resistivity at 300 K is ∼0.5 Ω cm). This, in turn, renders the material suitable for application as an electrode material in supercapacitors. This aspect has indeed been examined by the authors, but preliminary results indicate a rather low specific capacitance of ∼55 F g⁻¹. The composites are also magnetic, because of the presence of both Fe and Fe₃O₄. However, the saturation magnetization (∼30 emu g⁻¹) is not high because of the large proportion of carbon present. Despite the presence of CNTs, the mechanical properties of the composites are compromised by the porous morphology and the presence of a substantial proportion of graphitic carbon in addition to CNTs.

Conclusions

Low-pressure, inert-ambient CVD with Fe(acac)₃ as precursor results in a unique three-component composite of iron, iron oxide, and carbon. With Fe(acac)₃ as precursor, under appropriate conditions, MWCNTs are formed along with Fe₃O₄ – the first time (to our knowledge) that CNTs have been synthesized from a metal coordination complex (as opposed to hydrocarbons). Elemental iron, formed under the strongly reducing conditions of the inert gas ambient, apparently catalyzes the growth of CNTs. Indeed, some CNTs formed enclose iron particles within them. Formation of a composite coating of Fe/Fe₃O₄/C takes place, irrespective of the substrate used. In all the cases, the metallic iron formed is protected by the carbon that encloses it or surrounds it.

Acknowledgements

The authors acknowledge gratefully the Centre for Nano Science and Engineering (CeNSE) and the Advanced Facility for Microscopy and Microanalysis (AFMM) of the Indian Institute of Science, for supporting analyses by XPS, SEM, and TEM reported in this study. The work was supported by a grant under the NPMASS, Govt. of India.

References

1. E. T. Thostenson, Z. Ren and T.-W. Chou, Compos. Sci. Technol., 2001, 61, 1899–1912.
2. J. S. Sakamoto and B. Dunn, J. Electrochem. Soc., 2002, 149, A26–A30.
3. L. Yang, S. Cheng, Y. Ding, X. Zhu, Z. L. Wang and M. Liu, Nano Lett., 2011, 12, 321–325.
4. M. Zhi, C. Xiang, J. Li, M. Li and N. Wu, Nanoscale, 2013, 5, 72–88.
5. Q. Zhang, J.-Q. Huang, W.-Z. Qian, Y.-Y. Zhang and F. Wei, Small, 2013, 9, 1237–1265.
6. S. Bose, T. Kuila, A. K. Mishra, R. Rajasekar, N. H. Kim and J. H. Lee, J. Mater. Chem., 2012, 22, 767–784.
7. K. Woan, G. Pyrgiotakis and W. Sigmund, Adv. Mater., 2009, 21, 2233–2239.
8. M. Zhu and G. Diao, Nanoscale, 2011, 3, 2748–2767.
9. V. Subramanian, H. Zhu and B. Wei, Electrochem. Commun., 2006, 8, 827–832.
10. M. F. L. de Volder, S. H. Tawfick, R. H. Baughman and A. J. Hart, Science, 2013, 339, 535–539.
11. E. Flahaut, A. Peigney, C. Laurent, C. Marlière, F. Chastel and A. Rousset, Acta Mater., 2000, 48, 3803–3812.
12. Y. He, L. Huang, J.-S. Cai, X.-M. Zheng and S.-G. Sun, Electrochim. Acta, 2010, 55, 1140–1144.
13. D.-W. Kim, K.-Y. Rhee and S.-J. Park, J. Alloys Compd., 2012, 530, 6–10.
14. B. Gao, C. Peng, G. Z. Chen and G. Li Puma, Appl. Catal., B, 2008, 85, 17–23.
15. J. Li, S. Tang, L. Lu and H. C. Zeng, J. Am. Chem. Soc., 2007, 129, 9401–9409.
16. H. Zhang, G. Cao, Z. Wang, Y. Yang, Z. Shi and Z. Gu, Nano Lett., 2008, 8, 2664–2668.
17. S.-B. Ma, K.-Y. Ahn, E.-S. Lee, K.-H. Oh and K.-B. Kim, Carbon, 2007, 45, 375–382.
18. Y. Zhang, M. Xu, F. Wang, X. Song, Y. Wang and S. Yang, J. Phys. Chem. C, 2013, 117, 12346–12351.
19. J.-S. Ye, H. F. Cui, X. Liu, T. M. Lim, W.-D. Zhang and F.-S. Sheu, Small, 2005, 1, 560–565.
20. Y. Zhang, X. Sun, L. Pan, H. Li, Z. Sun, C. Sun and B. K. Tay, Solid State Ionics, 2009, 180, 1525–1528.
21. P. H. Jampani, K. Kadakia, D. H. Hong, R. Epur, J. A. Poston, A. Manivannan and P. N. Kumta, J. Electrochem. Soc., 2013, 160, A1118–A1127.
22. L. Marot, T. de los Arcos, A. M. Bünzli, C. Wäckerlin, R. Steiner, P. Oelhafen, E. Meyer, D. Mathys, P. Spätig and G. Covarel, Surf. Coat. Technol., 2012, 212, 223–228.
23. A. Varade, A. S. Shivashankar, S. Dhar and S. Sampath, US Pat., 8343572 B2, 2013.
24. A. Varade and S. A. Shivashankar, Carbon, 2011, 49, 1401–1407.
25. L. Machala, J. Tuček and R. Zbořil, Chem. Mater., 2011, 23, 3255–3272.
26. R. M. Cornell and U. Schwertmann, The Iron Oxides: Structure, Properties, Reactions, Occurrences and Uses, Wiley, 2003.
27. Z. Shao-Min, Z. Xing-Tang, G. He-Chun, Z. Bin, W. Zhi-Shen, D. Zu-Liang and W. Si-Xin, J. Phys.: Condens. Matter, 2008, 20, 075217.
28. S. Mathur, V. Sivakov, H. Shen, S. Barth, C. Cavelius, A. Nilsson and P. Kuhn, Thin Solid Films, 2006, 502, 88–93.
29. J. Bachmann, J. Jing, M. Knez, S. Barth, H. Shen, S. Mathur, U. Gösele and K. Nielsch, J. Am. Chem. Soc., 2007, 129, 9554–9555.
30. W. Jiang, Y. Zhou, Y. Zhang, S. Xuan and X. Gong, Dalton Trans., 2012, 41, 4594–4601.
31. T. Maruyama and Y. Shinyashiki, Thin Solid Films, 1998, 333, 203–206.
32. S. Dhara, G. L. Malhotra, A. C. Rastogi and B. K. Das, Thin Solid Films, 1992, 209, 116–121.
33. A. C. Rastogi, S. Dhara and B. K. Das, J. Electrochem. Soc., 1995, 142, 3148–3156.
34. K. Kuribayashi and R. Ueyama, Thin Solid Films, 1997, 295, 16–18.
35. C. N. R. Rao and A. Govindaraj, Acc. Chem. Res., 2002, 35, 998–1007.
36. R. Sen, A. Govindaraj and C. N. R. Rao, Chem. Phys. Lett., 1997, 267, 276–280.
37. N. S. Kim, Y. T. Lee, J. Park, J. B. Han, Y. S. Choi, S. Y. Choi, J. Choo and G. H. Lee, J. Phys. Chem. B, 2003, 107, 9249–9255.
38. J. Prasek, J. Drbohlavova, J. Chomoucka, J. Hubalek, O. Jasek, V. Adam and R. Kizek, J. Mater. Chem., 2011, 21, 15872–15884.
39. S. Dhar, P. Arod, K. V. L. V. Narayan Achari and S. A. Shivashankar, MRS Online Proc. Libr., 2015, 1752 DOI:10.1557/opl.2015.7.
40. X. Zhao, D. Xia and K. Zheng, ACS Appl. Mater. Interfaces, 2012, 4, 1350–1356.
41. Y. Li, Q. Meng, S.-M. Zhu, Z.-H. Sun, H. Yang, Z.-X. Chen, C.-L. Zhu, Z.-P. Guo and D. Zhang, Dalton Trans., 2015, 4594–4600.
42. J. Li, Z. Ai and L. Zhang, J. Hazard. Mater., 2009, 164, 18–25.
43. M. Busch, W. Schmidt, V. Migunov, A. Beckel, C. Notthoff, A. Kompch, U. Bergmann, M. Winterer and B. Atakan, Appl. Catal., B, 2014, 160–161, 641–650.
44. P. Bhunia, G. Kim, C. Baik and H. Lee, Chem. Commun., 2012, 48, 9888–9890.
45. F. Tuinstra and J. L. Koenig, J. Chem. Phys., 1970, 53, 1126–1130.
46. A. C. Ferrari and J. Robertson, Phys. Rev. B: Condens. Matter Mater. Phys., 2000, 61, 14095–14107.
47. P. K. Chu and L. Li, Mater. Chem. Phys., 2006, 96, 253–277.
48. R. Escribano, J. J. Sloan, N. Siddique, N. Sze and T. Dudev, Vib. Spectrosc., 2001, 26, 179–186.
49. Y. Wang, D. C. Alsmeyer and R. L. McCreery, Chem. Mater., 1990, 2, 557–563.
50. O. N. Shebanova and P. Lazor, J. Solid State Chem., 2003, 174, 424–430.
51. A. M. Jubb and H. C. Allen, ACS Appl. Mater. Interfaces, 2010, 2, 2804–2812.
52. I. Chourpa, L. Douziech-Eyrolles, L. Ngaboni-Okassa, J.-F. Fouquenet, S. Cohen-Jonathan, M. Souce, H. Marchais and P. Dubois, Analyst, 2005, 130, 1395–1403.
53. M. S. Dresselhaus, G. Dresselhaus, R. Saito and A. Jorio, Phys. Rep., 2005, 409, 47–99.
54. C. S. Kuivila, J. B. Butt and P. C. Stair, Appl. Surf. Sci., 1988, 32, 99–121.
55. N. S. McIntyre and D. G. Zetaruk, Anal. Chem., 1977, 49, 1521–1529.
56. T.-C. Lin, G. Seshadri and J. A. Kelber, Appl. Surf. Sci., 1997, 119, 83–92.
57. A. Schneider, Corros. Sci., 2002, 44, 2353–2365.
58. Q. Wei, E. Pippel, J. Woltersdorf and H. J. Grabke, Mater. Corros., 1999, 50, 628–633.
59. C. Giordano, A. Kraupner, I. Fleischer, C. Henrich, G. Klingelhofer and M. Antonietti, J. Mater. Chem., 2011, 21, 16963–16967.
60. R. M. Cornell and U. Schwertmann, in The Iron Oxides, Wiley-VCH Verlag GmbH & Co. KGaA, 2004, pp. 9–38,  DOI:10.1002/3527602097.ch2.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5ra07472j

---